Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
The nature of change in households
Three major determinants of household ability to adapt to change can be identified as a level of resources adequate to allow maximization of advantages and new options; perspectives, social stereotypes, and power structures that encourage rather than inhibit modifications in new circumstances; and equitable ways of sharing responsibility for basic social functions like reproduction and dependent care.
Form
The incidence of different forms of the household or primary living unit is determined by a combination of historical, cultural, and economic factors. Different forms may predominate at different periods. Household forms will themselves vary through time, a process which both influences and is influenced by its own strong cultural rationale, though this may or may not change at the same pace. However conspicuous any of these variations may be, their real significance lies in how they affect the ability of the household to provide for its members.
Four household forms represent trends at the end of the twentieth century. These are the multi-generational extended household, the nuclear household, the female-headed household, and the single-person household. Examining some of the effects of macro-level economic, political, cultural, and social influences on their internal organization, external relationships, and functioning can help to distinguish between positive and negative elements of change and indicate areas where intervention is needed.
The terms household and family are often used interchangeably. Indeed, the Universal Declaration of Human Rights speaks of the family as the fundamental unit of society. However, apart from the implication of kinship, there is no universally accepted definition of the family, and in practice attention often centres on the household. Households are usually families in the sense that most of their members are kin of some kind, but not all family members live in the same household and households may include people who are not family members. Here the focus is on the household as a unit of co-residence, and the linkages between separate households containing family members bound by reciprocal rights and obligations is considered as an external relationship.
Multi-generational or Extended Households
The multi-generational household containing parents, their children, and their children's children, as well as other relatives at each level, was once the norm in most countries. This applied particularly in rural areas, and this kind of household still predominates in regions of Africa, Asia, and Latin America. However, it is fast becoming less prevalent and the multi-generational households that do exist are often smaller than before. An individual is now much less likely to spend her or his entire lifetime in an extended household, though many people will still be part of one intermittently, at certain points in their life-cycle.
The introduction of the cash economy, the shift of the means and rewards of production from collective entities to the individual, and industrialization have changed the nature of the household and combined with the increasing separation of the place of work from the place of residence to make the multi-generational household less necessary as a unit of economic survival. The economic viability of large households was often eroded as the ratio of dependents increased with the rise in longevity and survival rates, while the resources that could be generated by domestic production declined in monetized societies. Agricultural development, which replaced traditional farming for internal consumption with cash-cropping, diminished the ability of rural households to meet the needs of their members. Some of them left to find ways of supporting themselves and other family members as urban or international migrants. The rapid urbanization - fed by large-scale rural-urban migration - that accompanied industrialization separated generations geographically. Changing perspectives emphasized the household as a unit of consumption rather than production. New educational and employment opportunities made households with even a single earner potentially viable.
The influence of socio-economic factors on the size and composition of the contemporary multi-generational household is seen in several of the country studies. Young married couples in China and Brazil remained in an extended household until they were able to accumulate sufficient resources to set up their own households. Most of them stayed in this conjugal nuclear household, but there were others whose need for the support of the original household sent them back into it. This sometimes happened when the first child arrived in two-earner nuclear families in Argentina. In Sri Lanka married daughters returned to their rural parents' households in times of illness and crisis, and that study refers to the conscious reformation of the extended family for convenience when young mothers were employed locally or in West Asia.
This episodic dependence on multi-generational households is also seen at another point in the life-cycle when economic and personal support is essential-old age. Elderly parents, especially widows, may move in with their children, forming a three-or even four-generational household. Variants of this phenomenon have been observed in developing and developed countries alike. A study of family support for the aged in Latin America came to the conclusion that these families might often "be held together less by ties of affection than by force of necessity." A profile drawn from a household survey in Brazil depicts an old person typical of the 56 per cent living in a multi-generational households as an illiterate widow, a relatively recent migrant from the countryside with few if any resources of her own, who was likely to be older and more incapacitated than members of her cohort who lived in other types of households. The house and household she lived in were likely to be very poor and her perceptions of her situation were frequently negative. Her counterparts who were better off materially and physically tended to live with a spouse (32 per cent) or alone (Ramos, 1990).
In Japan the migration of the elderly to live with younger generations of their family is a major influence on household size and shape: the increasing number and longevity of the older people there is, as in other countries, compensating for the smaller number of children in contemporary families. Fifty-nine per cent of the elderly live in multi-generational extended families. This group tends to be the oldest of the aged, most of whom have lost their spouse. Until they need more physical care, those old people who have the resources extend their independence by living close to, but not with, their families (Kong, 1990). This pattern has been observed in a number of industrialized countries.
The same factors that have reduced the prevalence of multi-generational households have produced a higher incidence of nuclear households consisting of a couple and their children. Declining fertility associated with family-planning and health services, and female access to health, educational, and employment opportunities, mean that these households have become smaller in most parts of the world. Most are headed by a man whose authority, control of resources, and position as the focus of attention and policies often remain unquestioned, although his income-generating role is increasingly shared with his partner.
Though many of the characteristics of a nuclear household are shaped by its economic environment, changing socio-cultural attitudes arc also important. A high value on privacy and individual autonomy make the conjugal nuclear household an option increasingly preferred even where extended households are still prevalent. Economic and demographic pressures may be responsible for the rising number of nuclear households bound by consensual rather than legal ties, but even the majority which arc based on formal marriage may be formed and dissolved more easily in societies increasingly tolerant of personal choice as a factor in both these events. Emotional or psychological motives may dominate in decisions about marriage and divorce in more cultures today, but the newly acquired freedom of choice may also be used to select a partner at least partly on socioeconomic grounds, as the country studies of Sri Lanka and China show. Similar reasons may outweigh personal dissatisfaction when divorce is considered.
In those societies where the dissolution and re-formation of families is easy, which often depends as much on economic possibilities as on legal and social acceptance, people may belong to different types of households at various points in their lives. They may move perhaps from their family of origin to a conjugal household, then back to the first, or into a single-person or sole-parent household until remarriage and a return to a conjugal household, the internal composition of which may be changed by the presence of offspring from previous unions. In other words, household types may be considered and treated in practical terms not so much as discrete forms, but as part of a continuum through the life-course.
Though single-person households have become more common as individuals are able to be economically independent, they are still relatively rare. However, the proportion of single-person households is rising as they become part of transitional stages. In industrialized countries young people tire of the prolonged period spent in their household of origin because of extended education and delayed marriage. If this education has given them sufficient earning power, they may leave to set up a single-person household. Nevertheless, studies in several countries where this has been a highly desirable goal suggests these and similarly motivated joint peer-group households are very sensitive to rising costs. In Australia and the United States, young people are likely to return to their households of origin as rents rise.
Most of them will eventually leave the household of origin again after a relatively brief time to set up their own conjugal households. Whilst these are likely to endure until the death of the partner in some cultures, in others there is a strong possibility that divorce or separation will dissolve the union. This will lead to the formation of single-person or single-parent (usually female-headed) households, or expanding nuclear households, at least temporarily. If the union is ended by the death of one of the partners, the other may remain in what has become a single-person household for as long as she (for the surviving partner is statistically likely to be the woman) is able. This widow is likely to have a depleted economic base and little way of supplementing it: her low educational level and intermittent workforce experience will prevent her increasing her income substantially, even if legal or cultural barriers do not stand in her way. If she is older, it is very unlikely that she will have her own employment-based pension: the ILO has estimated that only 6 per cent of the women in the world will be receiving such benefits in the year 2000. She may have some residual rights to her husband's pension, though this would apply to very few women in developing countries. As time goes on she may become physically as well as economically incapacitated and so seek the security of her family and become part of an extended multigenerational household; or she may resort to institutional care, if it is available.
Because more older people are women, this type of single-person household is likely to be a female-headed household. Sometimes, though rarely, an extended household may be headed by a woman, and somewhat more often a nuclear household may also be headed by a woman, if she is the major income-earner and recognized as such. However, the term female-headed household typically refers to one consisting of a woman and her dependents who are usually, though not always, her children. Because women's opportunities for remunerated work are likely to be constrained by age or family responsibilities, and their pay levels are in any case lower than men's, households headed by women are a major component of what has come to be called the feminization of poverty.
This has wide-ranging implications, for belonging to a female-headed household is now part of the life-course of an increasing number of women and children in every region. The proportion they make up of total households shows that they are a significant component, rather than a minor aberration of a nation's social organization. Conservative official estimates suggest that between 20 and 25 per cent of the rural households in the Caribbean and Latin America arc headed by women. The proportion is even higher in some of the countries in the region like Trinidad and Tobago, the Netherlands Antilles, Jamaica, and Cuba. In the last country, the figure has doubled from 14 to 28 per cent in less than three decades. In some of the countries in sub-Saharan Africa over 40 per cent of the households are headed by women, and the figure is over 30 per cent in a number of others. In Ghana the national figure rose from 26 to 32 per cent between 1960 and 1984, but a breakdown by region gives figures of 12 per cent in the northern part of the country, where traditional structures have remained more intact and widows and divorcees are still reincorporated into extended family households, and 41 per cent in the central region, where this support system does not operate (WISTAT. 1989; Calderón, 1990; Appiah, 1990).
Difficulties of definition affect analysis of this, the most rapidly growing, household form. When the women are widowed, divorced, or unequivocally bearing sole responsibility and authority in the household, their de jure status is clear. Others become de facto heads of households when war, migration, seasonal work, separation, or abandonment force them into this role regardless of their legal status. Their situation may thus be potentially permanent or temporary. They may be the sole supporter of their family or be partly maintained by remittances or other resource flows from a spouse or members of their extended, though not co-resident, family. They arc usually the mothers of the children in the household, but in some regions, the Caribbean for example, they may be their grandmothers. Female-headed households often contain fewer people than male-headed households, and they usually include fewer potential earners. The situation of women who bear the major responsibility for providing for households that include other adults (who may or may not contribute to its support) is particularly complex. Statisticians and demographers are not the only ones who have trouble identifying the head of household in this case: both male and female informants show a strong tendency to attribute the position to the male, regardless of his role in supporting the unit, and, as the women in the Kenyan study learned as their children grew up, adolescent dependent sons may automatically assume the title on the basis of their gender alone.
However, it is possible to make a number of valid generalizations about this heterogeneous group. Though it is not a new form, its extraordinary increase and spread to societies and cultures where it did not previously exist is directly attributable to macro-level development processes and stages. The shift from communal productivity based on land and the human resources of the extended family marginalized and impoverished many women and men. They needed money to survive in a cash economy. Lack of access to new resources like education, paid employment, or technology often led to urban or international migration in an attempt to meet family needs. Policies to attract labour to a particular sector drew more men away from their family households. Sometimes desperation in the face of a perceived inability to fulfil family responsibilities caused fathers to abandon their families. This became easier as greater mobility, as well as the pressures and opportunities of the cash economy, weakened traditional social controls. These pressures also eroded the kin or community support systems that had formerly protected women left alone for any reason. When social mores and moral values seemed confused, irrelevant, or too restrictive, sexual unions formed without the support of either the traditional or the new system left many women and children vulnerable if they collapsed. In more developed countries an element of choice was added to the economic pressures and the easing of social sanctions that helped to create so many female-headed households. Sometimes this took the form of a positive decision to assert ideological or financial independence. More often it represented an escape from a situation made intolerable by authoritarian or oppressive behaviour.
Many of these factors were also instrumental in variations in the shape and prevalence of other household forms. The growing number of female heads of households were hit by a congruence of them at a specific stage in their country's development and a particular point in their own life-course. They, like other women and men, need to have a level of economic and social competence sufficient to assure they can cope with changing circumstances. The female-headed household is a transitional or permanent base of growing numbers of women and it deserves the same kind of recognition and support as other types of household, especially as it has the same functions to perform. As will be shown below, it is in its ability to meet the needs of its members that differences may warrant differential treatment.
Internal Relationships
Attitudinal change has been a factor in all the contemporary variations in household form. While it is easy to observe these outward changes, variations in the internal relationships of households may be more difficult to recognize; they seem to come about more slowly, too. Practical recognition of changing roles and responsibilities within households is sometimes very slow indeed.
Change has tended to extend women's role by increasing their responsibility for providing household cash income. The migration of young Sri Lankan mothers to West Asia to earn money for their families is an extreme example of this, but rising costs and expectations, inflation, recession, and the harsh impact of many stabilization and structural adjustment policies mean that increasing numbers of mothers have to combine reproductive, remunerative, and domestic work. In Japan, the proportion of married women aged between 20 and 49 in the formal workforce rose from 13 per cent in 1963 to 40 per cent in 1989. In Australia only 18 per cent of all families with children, which represents 16 per cent of all households, depends on a sole income-earner. In Canada the figure is about 16 per cent (Ogawa and Ermish, 1990). In some societies, poor households may depend on the economic activities of their women and children for up to 50 per cent income. While more and more of them join the formal labour force, many others generate income in cash or in kind in the informal sector or by household production.
It might be logical to assume that as women extend their domestic responsibilities to include the provision of cash, other members of the household will share the household tasks that have been customarily allocated to them. Observation and the available data give little indication that this is happening. Nor do they suggest that changes in the control and management of economic resources within the household have come anywhere near to matching the magnitude of the shift in patterns of acquiring them. The persistence of behavioural patterns inappropriate in new situations can be traced to gender roles and power structures which appear highly resistant to change. It is very difficult to chart the socio-psychological aspects of domestic relationships, but it is essential to try to assess them because of their influence on household form and function.
Demographic trends are already causing changes in both the shape and internal relationships of many households. Lower fertility and greater longevity mean that the generational structure of households will be more like a beanpole than a pyramid in shape: there may be fewer people in every generation, but more generations. Changing attitudes to non-formal unions and to divorce and separation mean new combinations of blended families. In some cultures economic difficulties may cause a rise in the number of foster children. Later marriage and entry into the workforce and increased time spent in education mean the period in which young family members are financially dependent on their households of origin is prolonged.
The proportion of a woman's life devoted to child-rearing is declining as fertility decreases: it now occupies only about seven years of her life in an industrialized country and about 16-17 years in a developing country. Secondly, whilst the emotional and financial dependence of childhood may be extended, the years of physical dependence are not, and the emphasis on maternity as the focal point in a woman's life should diminish as she becomes more available for other productive roles. However, whilst women's traditional role as mothers may occupy a lesser proportion of their life-cycle, it may become transmuted into economic activity to cover the rising expectations or extended dependency of their children. Then, in their middle years, women may have to reassume a primary caring responsibility as the daughters or daughters-in-law of the longer-living older generation.
None of these family roles or relationships is new, but different emphases and time-spans may have significant effects on household organization. They may also be a major determinant of whether the cohesion of the domestic group is reinforced by shared affections, values, and goals, or undermined by conflict over these and the distribution of household resources and responsibilities. As the solidarity this cohesion represents is an important factor in the unit's ability to function to maximum advantage and to provide the most satisfactory base for its members' activities outside it, it is necessary to recognize potential of friction. For example, if adolescent and young adult offspring arc denied the input into decision-making or the emphasis on their interests and autonomy that they enjoy in the wider community, tensions are bound to arise. If old people are prematurely relegated to dependence and inactivity by social attitudes, legal constraints, and economic difficulties, their increasing need to share scarce resources may lead to resentful competition. If rhetorical respect for women and their rights as equal citizens is denied by custom and structures within the home, their potential is diminished. The gap between their expectations and prerogatives in their broader environment and their actual status in the domestic group has already made some young people and women of all ages reluctant to be part of the more conventional household forms.
A major cause of their dissatisfaction is socially reinforced systems of domestic authority based on male dominance in decision-making. Whilst their impact in certain areas like sexuality and contraception and labour-force participation is recognized, it pervades almost all spheres of domestic activity. It is difficult to measure this intangible power, but it can be traced through the allocation of household income, an area in which its force was previously misunderstood. Recent research has shown that while women may appear to be in charge of this, in fact their role is much more limited: they are most often managers of that portion of household income which the men who in tact control it allocate to them. The boundaries of women's power usually coincide with those of gender-based divisions of labour and interests which confine them to areas life food, child and dependent care, clothing, rent, and day-to-day expenses. In some areas like education the decisions may be taken by men, but the responsibility for paying for them may rest with women. Decisions about expenditure that is less routine or more external, like the maintenance of the dwelling, holidays, or the purchase of major items, are also more likely to be made by men. Several of the country studies below support the findings of surveys in developed countries that indicate that even when women themselves declare such decisions to be jointly taken, this represents a conscious or unconscious magnification of their input and influence.
Patterns in the control and management of household income seem to vary more between socio-economic strata than they do between cultures or countries. The strategies used by comparable (in relative terms) income groups in every region are remarkably similar. Women control and manage income in male-headed households (which most extended and nuclear households are assumed to be) only if they are the sole provider. Women who head single-person or single-parent households often cite complete control of resources as one advantage of their situation. As the proportion of the household income earned by women decreases, so their role reverts to management rather than control. This widely observed phenomenon relates less to economic rationality and competence than to male and female attitudes that see women's income-generation, whether in cash or kind, as merely a supplement to male earnings. Women's remunerated work is therefore perceived not as a primary economic function but as an extension of their social roles as helpmates to husbands and purveyors of family care.
The two points at which exceptions to this pattern are likely to occur throw further light on the broader question of the economic foundation of power structures. Women may also acquire total control in dual- or single-earner households if the income is so low that it is extremely hard to eke it out to cover basic needs. This pattern, which appears in studies of low-income sectors and families relying on social welfare benefits in industrialized countries, may also be seen as a way of extending women's responsibilities for the household's standard of living when they become too heavy for men. The increased domestic and income-generating activities of many women in developing countries, which constituted an invisible micro-level adjustment to recent international crises, are another facet of this trend.
Women are more likely (though by no means certain) to have more real control if they enjoy middle to high incomes and employment status which equals or exceeds their husbands'. Some research has detected an interesting tendency to underplay women's superior earning power in households where to acknowledge it would undermine the socially endorsed status of the male breadwinner. However, this is a rare problem. Quite apart from cultural perceptions which undervalue their economic input, whatever form it takes, most women do not have the opportunity to make equal cash contributions to their households. Lower levels of education and training, their concentration in poorly paid jobs or economic sectors, social stereotypes and other barriers to employment and promotion, and intermittent and atypical workforce participation - shaped, ironically enough, by the very family responsibilities they are trying to meet-mean there is no country in the world where women's waged income equals men's. In Chile, for example, the average adjusted female figure for the period 1960-1985 was no more than 68 per cent of male income. In 1985, it ranged between 53 and 84 per cent in five major cities in Latin America. In Switzerland there was an overall difference of approximately 30 per cent in 1986. Even when they worked in similar employment categories, there was a difference of some 20 per cent between what women and men earned (ECLAC, 1988; Swiss Federal Bureau of Statistics, 1989).
This unbalanced pattern of authority and control may have severe repercussions on household well-being. First, it runs counter to the conventional wisdom, now supported by research, that women are more likely to be aware of family needs and to give priority to expenditure on them. But their ability to do this will be constrained if they can only manage a limited amount of money which may be allocated with less sensitivity to family needs. A number of studies in industrialized countries have shown that the amount allocated by a man for general household expenditure often fails to keep pace with rising costs. Research in a range of countries suggests that men are more likely to devote money for personal consumption to leisure and greater or smaller luxuries, while women tend to spend more of any surplus or personal funds on their families.
Second, government policies which overlook this unbalanced control of incoming resources may miss their targets. Tax rebates, child allowances, and income maintenance systems which ease the father's financial situation or channel payments through him do not necessarily achieve their stated aim of helping households and families. The determination of mothers in industrialized countries to have family allowances paid directly to them reflects their practical need to control expenditure as much as the expressed desire for independence.
The same unequal division of power maintains inequitable allocations of domestic responsibilities. It is a circular process whereby the fact that women do undervalued domestic work reinforces stereotypes of male superiority. This situation persists in spite of the heavy burden of labour-force activities women have had to assume in addition to their non-market work and reproductive and caring functions. Socio-legal appreciation of women as equal participants in all spheres of activity and the declarations of many women and younger people of both genders that shared domestic responsibility is more appropriate in today's world make little practical difference. The data from the time-use surveys and qualitative information in the country studies are corroborated by a growing body of research in this field. Male attitudes towards women in Barbados typify widespread assumptions ("Certain things make for a man and certain things make for a woman"), in fields ranging from sexuality ("A woman is here for a man") to domestic labour ("Women should take their place and do the housework. I think that a woman's place is in the home") (Dan, 1986). Though they may be less blatant, these concepts underlie male and female attitudes in every region. They appear impervious to the pressures imposed by women's incorporation into the labour force.
It has been estimated that, while married, women whose productive work was confined to non-market domestic activity in Chile in the early 1980s devoted five to six hours a day to household work. Those who responded to rising male unemployment, lower real wages, and rising prices by taking on waged work still did an average of 37.9 hours' housework a week. A study which included six other countries in the region found no corresponding increase in male participation in domestic labour (ECLAC, 1988). Women agriculturists in Guyana spent three to five hours of their 15 1/2 -hour working day on household chores; their men helped with some of the farm work only if there were no older female children to do so. A study of Jamaican women factory workers suggested their household responsibilities were heavier in male-headed households, where the tasks were not shared, than in female-headed households, where other women and the children helped them (Bolles, 1986).
In 1987, a year-long survey of 1,668 households in the Netherlands concluded that the division of paid and unpaid labour, leisure activities, and child care was unequal. When female labour-force participation was limited by the presence of young children, 53 per cent of the men said domestic work should be shared. When both partners were in full-time employment, 84 per cent of the men agreed the housework should be shared. In fact, it devolved mainly on the women (Stoop and Oudhoff, 1989). This apparent example of attitudinal change preceding behavioural change may reflect the element of choice which men feel in this matter: they are willing to assist with cooking, washing up, or some of the more interesting areas of child care, for example, but the primary responsibility remains with women who have little choice but to do everything else. In Switzerland women devote an average of six times more time to household tasks; even those in full-time employment spend four times as much on it (Swiss Federal Bureau of Statistics, 1989).
Time budgets show many employed women must adopt the strategies of the women in the Kenyan study, who had to "make the day longer" by getting up earlier and going to bed later, denying themselves rest in order to get their household tasks done. A time-use survey of 12 countries as disparate as Belgium, Hungary, Peru, the United States, and the Soviet Union found men had an average of 34 hours free time a week; women in full-time household work had 32.6 and employed women had 14.5 (Waring, 1988). Many women in developing countries would share the Kenyan women's feeling that the concept of leisure time was irrelevant in their circumstances.
Just as the Kenyans did not have enough time to attend literacy classes or organize women's groups, so women everywhere arc precluded from other activities because domestic work absorbs so much of their time and energy. This is probably as strong a factor as outdated stereotyping and prejudice in their continuing low levels of representation in formal political and trade union activity. It is easier for women to participate in community level and non-government organizations whose structures may be more responsive to the demands of their household responsibilities, but their input is very much needed at the higher levels where the decisions on macro-policies are made.
The care of the young, sick, disabled, and old members of the household is part of the function of the basic social unit. However, because the primary responsibility for this is usually allocated to women, inequitable relationships within the household are accentuated and women, the traditional carers, are further denied time and opportunity. The extended household has advantages over those where there are fewer people to share this task, and fertility and female labour-force participation rates may be higher in them. But extended households are decreasing, even in developing countries, at the same time as the number of those needing care is increasing. While the number and proportion of infants and children needing care is falling, the obverse of the demographic transition, the ageing of the population, means the number and proportion of the elderly who may need care will continue to rise. This is demonstrated by dependency and caring ratios, which express the proportion of those who need care in relation to the number of potential care-givers. In 1950 in industrialized countries there were 22 old people for every 100 potential care-givers. The ratio had risen to 33 to 100 by 1985 and it is projected to be 55 to 100 by the year 2025. Though this dilemma was once thought to be specific to industrialized countries, it is now clear that the ratio will rise significantly in developing countries, from 14 dependents per 100 carers in 1950 to 25 per 100 in 2025 (UNDAW, 1988). Women there will have less support from shrinking extended families and governments which face fiercely competing demands on their resources, though they may continue to have the advantage of cultures in which old people can remain active and productive without the legal and social impediments that may impede them in more developed countries.
Women in all regions will be increasingly called on to make up the shortfall in the workforce due to the drop in the younger population. The burden of caring for dependents of all ages will have to be shared. The extent to which governments and institutions can participate in this will be severely limited by resources that are likely to fall as the tax base narrows. An obvious strategy will be to broaden the ranks of the carers by sharing the work among both genders.
In spite of its often-cited connection with the biology and psychology of women, the responsibility for taking care of the aged is in fact assigned on the basis of cultural expectations and personal availability. In 1987 a survey of primary caregivers in four localities in China showed three in every five carers were men. A typical carer was under 40, married, had completed secondary education, and was in full-time employment. Both the urban and rural carers were overwhelmingly classified in the middle-income range. Day-to-day physical and financial care were provided. "Duty towards the elderly" was their main reason for assuming this responsibility, and the majority (56 per cent) of the carers expected their sons to look after them in their old age. Only 21 per cent wanted their daughters to do this. In fact, more than two-thirds of the carers were the sons or daughters-in-law of those being looked after (Concepción, 1990). Expectations of care from sons and their wives may be an important factor in retaining a preference for bearing and educating boys, and decisions made on this basis at the micro-level may have a significant effect on the success of national policies on population, education, and support for the aged.
In 1985 the General Household Survey (GHS) of the United Kingdom was used to investigate dependent care. The results were surprising in that they indicated almost the same proportion of both sexes were involved: 15 per cent of the women and 12 per cent of the men in the 10,000 households surveyed were caring for someone who was sick, handicapped, or old. The dearth of earlier comparable data makes it difficult to know whether the GHS has illuminated a situation previously misunderstood or discovered a new trend. However, some of the data were not so surprising: 59 per cent of the women, but only 39 per cent of the men, bore the main responsibility for ongoing basic care. The male carers were more likely to undertake less constant or onerous tasks, a parallel situation to that revealed in studies of shared child care and housework. The GHS also shows that the nuclear household may still carry out functions formerly undertaken within the extended household: 11 per cent of the women in the sample and 8 per cent of the men cared for a dependent, usually a parent, who lived elsewhere (Green, 1989). The extension of a formerly internal activity into an intra-household function can be seen in geriatric care in both developed and industrialized countries, though like the reverse flow of help in cash or kind from the households of older to younger family members, this can only occur at a certain economic level.
New ways of identifying the patterns of authority, resource allocation, and domestic and caring responsibilities in households suggest they are not changing as rapidly as external circumstances would suggest or require. Deeply entrenched ideas about gender roles appear to be the major reason for this, though younger people are beginning to express attitudes more in keeping with current ideologies and demands. However, whether it relates to joint decision-making or shared responsibility for child care, the behaviour consistent with these attitudes is much more likely to occur when it is facilitated by the level of economic resources. Generally speaking, women now have an equitable part in financial decision-making and the power this confers only if they contribute as much or more income than their partners. Domestic tasks are more likely to be fairly apportioned when both parties enjoy equivalent educational and employment status, an effect which operates only when both are above the average attainment. Parents in dual-earner families do not usually take the maximum advantage of provisions designed to facilitate shared dependent care. It is almost always the mother who resorts to part-time or atypical work in an attempt to reconcile family and labour-force commitments. This increases women's financial dependence and reinforces patterns of domestic subordination, but as long as men's remuneration and promotion prospects are better, this option will be in the best material interests of the household. The issues of pay equity and equality of access to employment opportunity are therefore crucial to the establishment of more balanced intra-house-hold relationships.
External Relations
As the basic unit of social organization, the household both reflects and influences the larger units. Value systems are cultivated and transmitted through the household as a centre of socialization. Macro-level events and national policies have a conspicuous impact on household, but their direction and effectiveness is strongly influenced by household response to them, as in the fields of population and education. Some of the most significant macro-level developments - demographic transition or increasing female labour-force participation, for example-are in fact the outcome of the accumulated actions or reactions of households.
The web of mutually responsive linkages which bind the primary living unit and other social groupings will vary according to the status accorded to the form of the unit and to the individuals of whom it is composed. The key issues in examining these relationships might be summarized as matters of reciprocity of obligation. As individuals and households inevitably surrender more of their autonomy to larger socio-political groupings which are better able to meet certain needs, for example in health, education, or infrastructure, the ethical and practical imperatives for support systems for those who carry out the basic societal tasks of reproduction and care become stronger.
The household is also an economic entity, a centre of production and consumption, as well as the base from which the labour force originates and operates. Its patterns of consumption, saving, and investment can have a considerable impact on economic growth. Equity and efficiency suggest that social support measures to enable households to cope with change and contribute to development demand that responsibility for dependent care be shared within the household and between the household and the public and economic sector. This sharing should be based on capacity and need rather than on gender or preconception.
The comforting myth of the extended household as a source of adequate care for all family members overlooks both the economic limitations on this function in contemporary society and the fact that many who seek its coverage do so reluctantly, only because they lack the means to be more independent. Yet expectations may persist to such an extent that the need for alternative support systems is not recognized, much less addressed. The caring role of the traditional extended family was almost certainly more limited when high mortality rates and low life expectancy meant there were not as many people surviving to need care. In societies at all stages of development, with weak or strong generational links and ideals of support for the aged, economic constraints may make it very difficult for households to supply this now. Customs and formalized systems which formerly protected people at vulnerable stages may not operate in altered circumstances. For example, in some African cultures widowed daughters and their children were reabsorbed into the parental household. Today, the pressures of the cash economy and shrinking land resources may make this impractical. In other societies, polygynous families enabled a widow to reassert her claim to her former husband's support by remarrying into his family. A lack of the means to support newcomers may now deny women this avenue of survival.
On the other hand, flows of cash and various kinds of material and nonmaterial assistance between family members who reside in different households may characterize the external relations of many domestic groups, as functions formerly provided within a co-resident household are now spread between a number of nuclear households. Remittances and two-way flows of resources (cash in one direction, food or child care in the other) are well-recognized examples of this in developing countries. Studies suggest that intra-household systems of caring, for example for nearby elderly relatives or children, and regular or occasional material support may be more common than previously realized in more developed areas, too. Other intra-household support systems may be strengthening as non-kin and peer networks and community groupings replace some former family ties.
Greater sensitivity to and support for household responsibilities might be expected from the economic sector which benefits from households as the base of the labour force, and from women as workers, consumers, and managers of family income. Employers and unions are now more sensitive to workers' need to balance workforce and domestic responsibilities, but they teed to regard domestic responsibilities as women's responsibilities, an approach which can marginalize women, reinforce their role as primary carers, and deny men the possibility of increasing their share of these responsibilities. Some schemes do seek to cater for parents rather than for mothers, but, as noted above, as long as women's labour-force participation is less remunerative and less secure, response to them will inevitably be biased.
Women should be recognized as an intrinsic and valuable component of the labour force, rather than a supplementary reserve of low cost, transient workers. If pay equity can be achieved, the economic rationale underlying the allocation to women of dependent care responsibilities will be undercut. This in turn will allow mothers and fathers, sons and daughters alike to participate in dependent care. Economic restructuring and technological change provide an opportunity to stimulate human resource development through policies that broaden the options of male and female workers. Training that emphasizes broad-based and transferable skills for women whose employment may be interrupted by pregnancy and lactation, and systems that offer proportional protection and rights to men and women who choose atypical work to fit in with domestic responsibilities, are as important as the provision of childcare facilities.
In the broader community, approaches more sensitive to the reality of dual-earner households should mean reorganization of infrastructure, schedules, and facilities which pre-suppose that households have someone free to carry out a wide range of activities during conventional working hours. Sensitization to gender issues in access to income and to educational and political activities, as well as the transformation of obsolete stereotypes, are also essential to ensure that the relationship between the household and its wider environment is supportive. A broad perspective of how the needs and contribution of individuals and households vary at different points in the life-cycle would encourage this. For example, recognition of the fundamental socioeconomic value of women's reproductive and caring roles, and of the feet that old age is part of a universal process that need not be induced by social attitudes or arbitrary legislation, would ease the development of ways to prepare and cater for these stages. Community and state resources may be reduced or subject to competing priorities, and while social welfare systems might be necessary to cover particular stages or crises, it is important that social support systems enhance the flexibility and long-term capacity of individuals and households to function effectively.
Female heads of households with dependents usually face particular difficulties in providing care and earning income. Yet their need to do so may be greater: in the Caribbean and parts of Latin America, for example, many women head multigenerational households with high rates of dependency. In 1985 households headed by non-married mothers were the largest recipient group of welfare payments in the United States (Population Crisis Committee, 1985). A World Bank study of Ghana found that, even allowing for those who received remittances from absent husbands, women farmers who headed households had less than two-thirds of the income, in cash or kind, of male household heads, a situation replicated in many African countries (World Bank, 1989). All over the world female-headed households are overrepresented among the poorest of the poor.
This characteristic poverty stems from three main sources: pre-existing inequalities in women's access to resources like land, education, and services, which might have equipped them to cope better in the economic sector; the circumstances which have thrust their role as major provider on them; and reduced household capacity for income generation after this has occurred. The first and third elements of this sequence are symptoms of uneven or gender-biased development which disadvantages women in all regions, although their difficulties in providing for their families later will be more acute than those of women whose households have more than one worker or more alternatives for dependent care and domestic labour. While some specific measures - socio-legal recognition of their situation and rights and supportive interventions during crises, for example- are needed, the steps that would give the mothers more options and autonomy and reduce the risk of trapping the children in these households in a continuing cycle of deprivation are those which would help women generally gain an adequate level of economic and social competence. These must begin in childhood with fair access to nutrition, health services, and education. They should be part of a continuum which ensures that legal recognition of the rights of all women are accompanied by policies that enable women to take advantage of them. Equitable access to education and training, to credit and economic opportunity, to family-planning and health services, and to assistance with dependent care would benefit all women, and therefore prevent or alleviate many of the problems that presently disadvantage female-headed households.