For the last few decades the alternative of either social participation or social exclusion has dominated the discussion of welfare, particularly in Europe. What worries people about persistently high rates of unemployment, for example, is not just that the unemployed may be poor but that they are likely to be ‘socially excluded’. Indeed discussion of ‘poverty’, in the industrialised countries of Western Europe has been largely supplanted by discussion of ‘social exclusion’. A concern with ‘exclusion’ from the benefits of citizenship in the welfare state lies at the heart of much of literature about social policy. In the European capitals of administration, economic planning, and labour advocacy, in Brussels, Paris and Geneva, the presses pump out weighty volumes of research on social exclusion (c.f. European Observatory on National Policies to Combat Social Exclusion; 1993; Scherer et al, 1998; Rodgers et al, 1995; Silver,1994; Silver and Wilkinson, 1995). In the U.K. the Blair government has even set up a cross-departmental Social Exclusion Unit to address this social evil (Mandelson, 1997).

The Frenchman René Lenoir is generally credited with having coined the term in 1974 (de Haan, 1998). He used the term to describe the ten per cent of the French population then excluded from employment-based, social security systems. In Lenoir’s day the excluded population included the disabled, suicidal people, the aged, abused children, substance abusers and so on. Since that time the term has been extended to cover the unemployed, migrants and minorities in France. In Britain there has been a clear continuity between the tradition of studying relative deprivation, poverty and social exclusion. Townsend’s famous relative deprivation scale is easily thought of as measuring how poverty leads to social exclusion. People were poor in Townsend's words;

The opposite of social exclusion is social participation, or integration in the activities customary in that society (de Haan, 1998: 12). As Atkinson points out social exclusion is a relative concept and ‘we cannot judge whether or not a person is socially excluded by looking at his or her circumstances in isolation’. ‘People are excluded from a particular society: it refers to a particular place and time’ (1998: 7).

Social exclusion is a multi-dimensional concept embracing economic, social and political deprivations (de Haan, 1998: 12-13). Exclusion from the labour market is only ‘one face of social exclusion…people may face exclusion in other parts of their lives, notably in the domain of consumption’ (Atkinson, 1998: 13). This kind of analysis goes beyond resource allocation and includes social identity, culture, agency and ultimately power relations.

The emphasis in the analysis is on agency and process. Social exclusion is always some action, that through a particular set of institutional arrangements, one social grouping does to another.

Moreover, these analyses usally refer to damaged prospects in the future, that temporary disadvantage may become permanent disadvantage. Work in this tradition has characteristically concentrated on dynamics. ‘People are excluded’, says Atkinson, ‘not just because they are currently without a job but because they [and their children] have little prospects for the future’ (1998:8). In this respect the concept of social exclusion comes close to the concept of vulnerability. Vulnerability is seen by some as that aspect of poverty that results in ‘insecurity, defencelessness, and exposure to risks and shock’ (de Haan, 1998:15).

Two immediate consequences flow from adopting the social exclusion framework — the first is a redefinition of the boundary between luxuries and necessities and the second is some consideration of the long term or life course effects of social exclusion. Let us examine each of these topics in more detail.

Let us turn first to a consideration of the distinction between luxury and need. For much of this century it has been customary to regard leisure expenditure as unnecessary expenditure. This is party because the tradition in studies of poverty has been to set such an austere standard that no one could challenge the standard as being too high. Rowntree’s pioneering study of poverty in York, late last century, specifically excluded leisure. In an effort to show that his budget was an irreducible minimum, Rowntree pointed out that:

Similarly, in 1993 Renwick’s and Bergmann’s ‘Basic Needs Budget’, which the authors see as extending the narrow approach of the current official U.S. poverty benchmark, covers seven area of expenditure — food, housing, health care, transportation, clothing, care and personal care/miscellaneous. Renwick and Bergmann note that the official standard includes allowances for ‘reading materials, recreational expenses, educational costs, alcoholic beverages, and miscellaneous expenses…. [t]he basic Needs Budget has no allowances for any of these items’ (1993, 10).

However, if social inclusion (participation) is to be encouraged and social exclusion avoided, then people must achieve a particular standard of living, relative to prevailing conditions in that community. Moreover the concept of social exclusion, as we have seen, covers all relevant activities which affect participation, pays special attention to effects on prospects (especially of children) and draws our attention to the capacity of others to limit one’s own capability to alter these circumstances. If the community standard is that people have access to leisure activities, then it follows from the social exclusion framework that people need recreation or leisure.

Recently the Social Policy Research Centre developed a set of indicative budget standards for Australia. Developing a budget standard involves describing a basket of goods (a pattern of consumption expenditure) that would ensure a particular standard of living, and estimating the cost of that basket at current market prices. In addressing this task Social Policy Research Centre relied on two methods of gauging community standards. The first method involved conducting a series of specially convened focus groups in urban, regional and rural communities. The second method was to rely upon the rule of thumb that if more than fifty per cent of Australians consumed the item in question then it formed part of the moderate (or ‘modest-but-adequate’) standard of living, and if more then seventy-five per cent of Australian consumed the item it should be part of even a frugal (or ‘low cost’) standard (Saunders et al, 1998: 66-73). Using the 50/75 rule is supported by long tradition of study into the social standards of deprivation (Mack and Lansley,1985).

Focus group responses were typically based on social exclusion criteria, and supported the idea that for the promotion of self-esteem, and to avoid damaging the prospects of children, the ability to sustain some personal care and leisure consumption was significant (Saunders et al, 1998: 523-572). Practically everyone in these groups thought television was basic item. It was often suggested owning a television was a means of saving the costs of more expensive out-of-home entertainment. Some access to the news via newspapers and support for hobby activities through magazines was considered important. Participants in all focus groups believed that the ability to purchase alcohol was a social necessity, even the income-constrained sole parents were reluctant to suggest that one could manage without alcohol. Consumption of alcohol was explicitly linked to the capacity to participate in a community form of social life. A regular hair cut, every six to eight weeks, was considered a basic entitlement and having to do without this damaged people’s self-esteem.

Most focus group participants thought a childhood without toys was a deprived childhood. Some even argued for the necessity of a home computer, so that children would not be educationally disadvantaged. Participants felt that ownership of private cars was most easily justified on the basis of parenthood. Australian parents need a car, respondents argued, if their children were not to be excluded from participating in key community activities — sport, ‘sleeping over’ with friends, children’s parties, etc.

Furthermore the focus groups and Atkinson follow the same logic in seeing the telephone as symbolic of the issue of social exclusion. The responses below are typical of those given by focus group participants

‘A person unable to afford a telephone finds it difficult to participate in a society where the majority have telephones’ says Atkinson (1998:14). ‘Children are not invited out to play, because neighbours no longer call round — they call up. Letters do not allow the same contact to be kept with relatives who have moved away. A person applying for a job may not be called for an interview since he or she cannot be contacted directly. ‘

Leisure is a peculiar thing. It is not itself a commodity, although as we has seen participation in leisure activities may require the consumption of commodities. Leisure also requires time. Indeed, the very definition of leisure is based on time. In popular discourse leisure is conceived as time at one’s own disposal. The concept of leisure is usually defined by contrast with constrained activities. In labour economics leisure is treated as the opposite of paid work. Leisure is often thought of as a residual. It is the ‘free’ time that remains after maintaining one's body in a healthy and socially acceptable state, contracting time to the market, and meeting domestic and family responsibilities. From the standpoint of social exclusion one can be ‘shut out of leisure’ by a lack of money or a lack of time. This is illustrated in Figure 1 below.

Those who are comparatively rich in both time and money, for example, wealthy retirees would fall into Sector I in figure above. For many this represents the pinnacle of social advancement — a society that affords its member both the financial resources and the time to enjoy these benefits,

Highly paid but overworked executives might be examples of that category of people in Sector 2, that is people who are relatively money rich but time poor. There are some industrial campaigns around the globe seeking to prevent working hours from putting any employee in this situation (Campbell, 1997 and Buchanan and Bearfield, 1997).

The unemployed are typically those that fall into Sector 3, combining a low capacity to the spend on leisure with an extraordinary quantity of ‘free time’ available to them. Combating social exclusion implies guaranteeing people in this situation sufficient income to permit enough leisure to satisfy the criterion of social participation.

Those with a low amount of free time available to commit to leisure and who have little capacity to spend on leisure goods and services, Sector 4 in Figure 1, are doubly excluded. Single parents fall into this doubly disadvantaged category. A budget standard aimed at avoiding the possibility of social exclusion would need to support not only leisure consumption but some measures to create more time for leisure, for example by offering carer’s relief.

As with all expenditure, there is a powerful relationship between income and leisure expenditure. By analysing the broad category of leisure expenditure in the 1993 Household Expenditure Survey (ABS, 1996) we can study both the strength of this relationship and the factors that affect it. In raw terms household leisure expenditure increases by $13 per week for each $100 dollars of extra (equivalent) disposable household income. A household on a median income, in the sense used here, spent $45.09 per week on leisure in 1993/4.

The relationship between equivalent income and leisure expenditure is shown in Figure 2. In this figure, equivalent household disposable income has been arranged from lowest to highest in deciles (bands of ten per cent). Apart from the lowest (first) decile, the expenditure on leisure rises consistently with increasing income. The aberrant leisure expenditure of the first decile may result from the low and fluctuating reported incomes of the self-employed.

The association between income and leisure expenditure remains powerfully significant even after controlling for life course stage and employment. The multiple regression analysis in Appendix A explores the effect of income and these other variables on leisure expenditure. Most of the major influences on leisure expenditure, independent of income, are the result of life course transitions. These are summarised in Figure 3

At the time of the most recent expenditure survey (1993/4), households without children (where the reference person is below the age of sixty) spend $57.55 per week on leisure goods and services when income is held constant at the median. The presence of pre-school age children reduces household leisure expenditure by $16.56 even when income and employment hours are held constant. Having a youngest child between five and twelve years of age has no significant effect on leisure expenditure, and this remains at the same level as households without children. When the youngest child is thirteen or fourteen there is an unexpected contraction of household leisure expenditure of $15.93. Children in their middle teens and young adult dramatic ally increase household expenditure on leisure activities above the level of households without children, by $15.16 and $34.03 respectively. Sole parent households spend $6.63 less on weekly leisure activities when the age of children is held constant. Households where the reference person is approaching age of retirement also spend less on leisure (-$9.07).

Surprisingly, independent of income and life course stage, hours of employment (for the reference person) have no measurable influence on household leisure expenditure. The presence of an employed spouse does make a significant difference, adding $7.76 to weekly household leisure expenditure. The idea that a spouse’s employment status affects patterns of consumption has been documented by Phipps and Burton (1992) and has been the basis of much of the discussion about marital bargaining in the literature (Lundberg and Pollak, 1996).

Reviewing these findings it seems reasonable to conclude that the major element leading to social exclusion from participation in leisure consumption is disposable income. The influence of employment and life course stages on the consumption of leisure goods and services, independent of income, is comparatively minor. Money is the key to this dimension of leisure participation. However, money is a necessary but not yet a sufficient basis for social participation in these activities, the other necessary element is free time.

An estimate of the amount of leisure time available to individuals was derived from a regression analysis of the 1992 Time Use Survey (ABS, 1993). The results of this regression analysis are set out in Appendix B.

The information on the duration of activities in the Time Use Survey is collected by the diary method. Thirty years of research has shown that the highest validity and reliability in the measurement of time spent in all activities is achieved by using time diaries, which are now used around the world (Baxter and Bittman, 1995; Goldschmidt-Clermont and Pagnossin-Aligisakis, 1995; Niemi, 1993; Pallié, 1993; Robinson and Gershuny, 1994; Robinson, and Godley, 1997).

Since individuals complete the diaries, it is possible to conduct a separate analysis for each gender. In this way it possible to peer into the ‘black box’ of household distribution of leisure time in a way that is not yet possible with data on leisure expenditure.

By contrast with access to leisure goods and service, time use surveys show income plays an insignificant role in the distribution of leisure time. The amount of leisure time available depends significantly on the individual’s sex, employment status, age, and family circumstances. Family characteristics include the presence of a spouse, age of spouse, employment status of the spouse and the presence of children of various ages.

Everything else being equal, men have 3 hours 17 minutes more weekly hours of leisure than women. Although the gender gap in leisure becomes narrower if gardening is considered ‘unpaid work’ rather than ‘leisure’, there is a persistent pattern of relative disadvantage for women evident in the results of the analysis. Whenever the influence of factors that increase leisure is considered, the gain is greater for men than women. And whenever the factors that reduce or constrain leisure are considered, their effects are less severe for men.

As might be expected, employment status has greatest influence of the amount of available time for leisure participation. Men who are unemployed or who are not in the labour force have between 17.2 and 19.1 hours a week additional leisure time available. For women, being unemployed or not in the labour force adds between 12.8 to 14.7 hours a week to the time available for leisure. The ‘leisure bonus’ associated with part-time work is substantially greater for men (8.3 hours per week) than women (6.5 hours per week). For both men and women, nothing short of having their spouse completely withdrawing from the labour force significantly adds to their own leisure time. In this instance, the effect of their partner’s withdrawal from the labour force, while beneficial for both sexes, is of greater benefit to women, a gender difference of 2.6 hours per week.

Parenthood profoundly reduces the amount of leisure time available to both men and women. Generally the impact is greatest when the child is very young and this impact diminishes as the child matures. Motherhood reduces leisure possibilities more seriously than fatherhood. Figure 4 summarises how the age of the youngest child affects parent’s leisure time, when parents are assumed to be aged 35 years, employed full-time and have a partner who is employed full-time.

Whereas men with no children have 17 per cent more leisure time than comparable women, fathers of children pre-school age (two years of age) have 56 per cent more leisure than comparable mothers. Once the youngest child is aged between two and nine years fathers have about a third more leisure time than mothers do, but once the youngest child reaches its teens the father’s advantage falls to a quarter.

Among men, when all other factors are held constant, leisure time increases slowly with age, while the age of their spouse has no significant effect upon their weekly hours of leisure. The estimated effect on the leisure a sixty-five year old man, compared to a twenty-five year old man, would be a net gain of 2 hours extra leisure per week, over and above the effect of retiring from the workforce. This holds true, whether or not the man is married. However, among women the obverse is true. Although women’s leisure time increases with advancing age, at a faster rate than for men, this is almost completely offset by a significant decrease their leisure as their spouse ages. Thus the net effect of reaching sixty five years for a single woman is a 6 hours per week increase in leisure time, over and above the effects of retirement. For otherwise comparable, married woman the net effect is a negligible increase in leisure time.

Leisure has a social patterning by weekday/weekend. Weekends are the time for extra leisure, which peaks on Sunday. All other things being equal, compared to women, men also enjoy more additional leisure on weekends. Starting from a higher base, compared to women, men enjoy both larger absolute amounts of leisure and a higher proportional increase on both Saturdays and Sundays.

Up to this point I have presented separately each of the factors affecting the distribution of leisure time. In reality, however, these effects are cumulative. A thirty-five year old mother of a pre-schooler, for example, is by definition female, has a child below the age of 5 years, may have a partner five years older than herself and be in full time employment. The predicted leisure time for this mother would be 15 hours per week. On the other hand, a seventy-year-old retired male could expectant to spend 57 hours 37 minutes a week in leisure activities.

A commonly employed standard used to benchmark (income) poverty internationally is fifty per cent of the median. Households with income that is lower than fifty per cent of the median are judged to be ‘poor’. Applying an analogous standard (fifty per cent of the median leisure time) to the basic household types used in the Budget Standards Project we can get some idea of what social situation produces the most severe kinds of time poverty. Half-median leisure time for the whole adult population is 19 hours 15 minutes per week.

The major causes of leisure time poverty are gender, family responsibilities and longer hours of work. Therefore those most at risk of time poverty have combinations of disadvantageous characteristics, this is shown in Table 1.

Full-time employed mothers of very young children are most at risk. This risk is greatest when they have a partner who also works full-time. Indeed mothers in a couple where both adults work full-time suffer from the most severe time poverty as long as their children are aged younger than 15 years, although it becomes a less severe time poverty as the child matures.

The data indicates that there are three things that might relieve this situation. The first path — part-time work — is commonly chosen by families and, according to Wolcott and Glezer (1995), the path heavily preferred by Australian women. Reducing weekly hours of work raises all women above the leisure-time poverty line, except for women with a youngest child below the age of two years and a partner in full-time work.

The other, less travelled path, where mothers continue to be employed full-time is to have a partner who is not in the labour force. According to the predictions of the regression model developed in this paper, the presence of stay at home fathers would raise the leisure time of their full-time working partners above the poverty line.

A third path, is divorce. Compared with their partnered counterparts in similar employment situations, lone mothers have significantly more leisure time. Lone mothers are not exposed to time poverty. Moreover, although fathers are relatively deprived of leisure, in comparison to men without children, they do not fall below the time-poverty line, even when their children are very young. It is possible that dissatisfaction with this situation is a motive for divorce. This finding draws attention to the differential impact of family responsibilities according to gender.

Inadequate leisure opportunities lead to social exclusion. Access to leisure is dependent on both time and money. Neither time nor money by itself is a sufficient condition for leisure. After studying patterns of distribution of both leisure consumption and time available for leisure activities, we might propose some tentative guidelines for policies aimed at addressing this form of social exclusion.

Leisure consumption opportunities are related powerfully to income. Few other factors have any statistically significant effect on the consumption of leisure goods and services. The influence of factors other than income is small. It follows that an adequate income security policy would ensure that households had the capacity to purchase sufficient leisure goods and services. Given an adequate income all households would be lifted above the leisure consumption poverty line, just as a rising tide lifts all boats. To date the two obstacles to formulating this policy have been the failure to acknowledge leisure needs and the lack of any standard by which to judge whether leisure expenditure is adequate. The neglect of leisure needs seems strongest in Anglo-Saxon societies with their history of dividing the destitute into the ‘deserving’ and the ‘undeserving’ poor. This tradition makes arguing for leisure needs politically hazardous and many, seeking to achieve some alleviation of poverty, specifically excluded any suggestion for the support of leisure. On the other hand, the very lack of tradition is a problem when it comes to estimating the money costs of leisure. Fortunately, the Budget Standard developed by the Social Policy Research Centre seeks explicitly to determine a standard of adequacy. The basis of this determination is set out in detail in the official publication (Saunders, et al, 1998).

The issue of the time costs of leisure is more complex and permits only a multi-pronged approach. As has been shown, leisure time poverty has three main causes - longer hours of work, family responsibilities and gender.

There is now a revitalisation of the industrial politics of working hours. Most specialists agree that standard working hours have been eroded by the casualisation of the workforce, the popularity of part-time work, the growth of the ‘24-hour society’ in retailing, early retirement, and so on (Donaldson, 1996). These discussions are most advanced in Europe, where ‘overwork’ and unemployment are seen as the two faces of the coin of maldistributed working hours (Buchannan and Bearfield, 1997). A re-regulation of working hours could address the problem of time poverty, regardless of its success in reducing unemployment. A central consideration is the situation of working parents. The ACTU has recently signalled its intention to campaign on issues related to working hours.

However, there has been far less attention paid to the time costs of parenthood, especially in the sense of social exclusion. With the birth rate falling across the advanced industrial world, renewed attention is being paid to the high costs of children (McDonald, 1997). The total fertility rate in Australia has fallen from close to six live births per woman in the middle of last century, to less than two at the close of this century (McDonald, 1995). In some countries fertility has fallen to levels that threatens something approaching extinction of a whole people. McDonald notes that ‘in a stable population with the level of fertility applying in 1995 in Italy, population size would drop in just 100 years to 14 per cent of its initial level’ (1997: 2). In explaining how advanced industrial societies reduced their fertility to such a low rate, Caldwell (1976) argued that this ‘demographic transition’ followed the reversal of the flow of wealth between the generations. In centuries gone by, having children had been a way of parents securing their economic future, in contemporary society children became a strain on their parents economic resources.

Although there is a voluminous literature on how to measure the ‘true’ costs of children (see Nelson, 1996 for a review of this literature), little attention has been given to time costs of children. Even less attention has been directed at the leisure disutility of children. Compared to those without children, parents enjoy less hours of leisure time. Moreover, some research I am currently undertaking with Judy Wajcman of the Australian National University, suggests that what leisure parents enjoy might be of a lower quality. Parent’s leisure is significantly more likely to be fragmented, to be combined with unpaid work activities, and parents are less likely to able to follow adult leisure pursuits.

What policy instruments could address the leisure gap between parents and non-parents? Parental leave is one method of freeing parent’s time by reducing paid work commitments. This could take the form of maternity, paternity leave or both. Paid parental leave ensures some level of redistribution of both time and money to parents. Some countries, especially in Scandinavia, have generous forms of parental leave, with a long duration (12 months, in some cases), high replacement rates (70-90% of earnings) and continuing entitlements to shorter working days for pre-school-age children (Leira, 1992). The data suggests that the ability of either parent to reduce hours of employment makes a significant difference. Child care subsidies are another instrument which could be effective in this context, especially if non-work related needs for child care were treated as legitimate (Bergmann, 1997). Guidelines for what would constitute an adequate level of ‘child carer’s relief’ would need to be developed. The Budget Standards employed a minimum based on sufficient time for physical exercise as a normative standard (Saunders, et al, 1998) and an alternative half-median standard has been proposed in this paper. Either standard could become a the basis for policy discussion.

Finally but unavoidably, thought must be given to the relationship between gender, work and family. The data presented here suggest that only women are at risk of time- poverty. The indications are that this results from the assignment of domestic and family responsibility by gender. In 1956 Myrdal and Klein first argued that there was a social revolution under way that it was now possible ‘for women to envisage the idea of work outside the home as a career for life without any feeling of self-denial or resignation’ (1968, p.8-9). Ever since this time there have been two contrasting predictions about how contemporary society will adjust to this change. Some have argued that an independent income will so strengthen women’s domestic bargaining power that men and women will share of housework and child care equally. The alternative has been to argue that it will prove difficult to change the assignment of family responsibilities by gender, and women will be obliged to assume a ‘dual burden’. In a pattern called ‘pseudomutuality’, the idea of the egalitarian family has been enthusiastically adopted but unfortunately without any actual reassignment of tasks (Bittman and Pixley, 1997:145-171). Men’s behaviour, as distinct from their beliefs, has proved difficult to alter. At the same time, women have been adapting to the new situation more quickly, seeking to vary their hours of employment and making adjustments to reduce the time spent in many domestic tasks. In this process the growth of institutions beyond the home which replace the domestic labour of wives has played a crucial role (Bittman, 1998). The institutions which give greatest leverage in combating gender equity, arising from the domestic division of labour, also turn out be the same institutions that most alleviate the leisure-time poverty of parents. An expansion of these services would be a step in ameliorating the social exclusion from leisure.

References

Apps, P. and Rees, R. (1995) ‘Household Production, the Costs of Children and Equivalence Scales’ Working Papers in Economics and Econometrics No.285 Canberra: Australian National University

Atkinson, A.B. (1998) ‘Social Exclusion, Poverty and Unemployment’ in A.B. Atkinson and J.Hills (eds) Exclusion, Employment and Opportunity, CASE/4 London: Centre for the Analysis of Social Exclusion, London School of Economics

Australian Bureau of Statistics (1996) Household Expenditure Survey, Australia: Household Characteristics Cat. No. 6531.0, Canberra

Baxter, J. and Bittman, M. (1995) ‘Measuring Time Spent in Housework: A Comparison of Two Approaches’ The Australian Journal of Social Research 1: 21–46.

Becker, G.S. (1965) ‘A Theory of the Allocation of Time The Economic Journal 75: 493-517

Bergmann, B. (1997) Saving Our Children from Poverty: What the United States Can Learn from France New York: Sage.

Bittman, (1998) ‘Changing Family Responsibilities: The Role of Social Attitudes, Markets and the State’ Family Matters, 50: 31-37.

Bittman, M. and Goodin, R.E. (1998) ‘An Equivalence Scale for Time’ SPRC Discussion Paper No.85 Kensington: Social Policy Research Centre, University of New South Wales

Bittman, M. and Pixley, J. (1997) The Double Life of the Family: Myth, Hope and Experience Sydney: Allen & Unwin.

Buchanan, J. and Bearfield, S. (1997) Reforming Working Time: Alternatives to Unemployment, Causalisation and Excessive Hours Melbourne: Brotherhood of St Laurence

Caldwell, J.C. (1976) ‘Toward a Restatement of Demographic Transition Theory’ Population and Development Review 22: 729-739.

Campbell, I. (1997) ‘The Challenge of Increased Precarious Employment’ Just Policy No11, pp. 4-20

de Haan, A. (1998) ‘"Social Exclusion": An Alternative Concept for the Study of Deprivation?’ IDS Bulletin 29: 10-49

Donaldson, M. (1996) Taking our Time: Remaking the Temporal Order Nedlands W.A.: University of Western Australia Press.

Goldschmidt-Clermont, L. and Pagnossin-Aligisakis, E. (1995) ‘Measures of Unrecorded Economic Activities in Fourteen Countries’ Occasional Paper No. 20 New York: United Nations Human Development Report Office.

European Observatory on National Policies to Combat Social Exclusion (1993) Social Services and Social Exclusion Lille: European Economic Interest Group ‘Animation and Research’

Leira, A. (1992) Welfare States and Working Mothers: The Scandinavian Experience Cambridge: Cambridge University Press

Linder, S. (1970) The Harried Leisure Class New York: Columbia University Press

Lundberg, S. and Pollak, R.A. (1996) ‘Bargaining and Distribution in Marriage’ Journal of Economic Perspectives 10: 139-158

Mack, J. and Lansley, S. (1985) Poor Britain London: George Allen & Unwin

Mandelson, P. (1997) Labour’s Next Steps: Tackling Social Exclusion Fabian Pamphlet 581, London

McDonald, P. (1997) ‘Gender Equity, Social Institutions and the Future of Fertility’ Working Papers in Demography No. 69 Canberra: Demography Program, Research School of Social Sciences, The Australian National University.

McDonald, P. (1995) 1995 Families in Australia: A Socio-Demographic Perspective Melbourne: Australian Institute of Family Studies

Myrdal, A. and Klein, V. (1968) Women’s Two Roles: Home and Work London: Routledge & Kegan Paul

Nelson, J. (1996) Feminism, Objectivity and Economics London: Routledge

Niemi, I. (1993) ‘Systematic Error in Behavioural Measurement: Comparing Results from Interview and Time Budget Studies’ Social Indicators Research 30: 229–244.

Paillé, B. (1994) ‘Estimating the Volume of Unpaid Activities in Canada, 1992: An Evaluation of Data from the General Social Survey’ Working Paper No. 10. Ottawa: Statistics Canada.

Phipps, S. and Burton, P. (1992) ‘What’s Mine is Yours? The influence of Male and Female Incomes on Patterns of Household Expenditure’ Working Paper 92-12, Department of Economics Dalhousie University.

Renwick, T. J. and Bergmann, B. R. (1993) ‘A Budget Based Definition of Poverty with Application to Single-Parent Families’ Journal of Human Resources 28:1-24

Robinson, J.P. and J. Gershuny (1994) ‘Measuring hours of paid work: time-diary vs. estimate questions’, Bulletin of Labour Statistics, xi-xvii, International Labour Office, Geneva.

Robinson, J.P. and G. Godbey (1997) Time for Life: The Surprising Ways Americans Use Their Time, Pennsylvania State University Press, University Park.

Rodgers, G., Gore, C. and Figueiredo, J.B. (eds) (1995) Social Exclusion: Rhetoric, Reality, Response Geneva: International Labour Organisation.

Rowntree, B.S. (1901) Poverty: A Study of Town Life London: Macmillan

Saunders, P., Chalmers, J., McHugh, M., Bittman, Bradbury, B. (1998) Development of Indicative Budget Standards for Australia Policy Research Paper Number 74 Department of Social Security, Canberra

Scherer, P., Hennessy, P. and Pearson, M. (1998) The Battle against Exclusion: Social Assistance in Australia, Finland, Sweden and the United Kingdom Paris: Organisation for Economic Co-operation and Development

Silver, H. (1994) ‘Social Exclusion and Social Solidarity: Three Paradigms’ IILS Discussion Paper Series NO 69 Geneva: International Labour Organisation

Silver, H. and Wilkinson, F. (1995) ‘Policies to Combat Social Exclusion: A French-British Comparison’ IILS Discussion Paper Series NO 83 Geneva: International Labour Organisation

Townsend, (1979) Poverty in the United Kingdom. A study of Household Resources and Living Standards Harmondsworth: Penguin

Wolcott, I. and Glezer, H. (1995) Work and Family Life: Achieving Integration Melbourne: Australian Institute of Family Studies

Source: ABS 1993/4 HES

	B	SE B	Beta	T	Sig T
Age	0.0512	0.0243	0.0342	2.1030	0.0355
Spouse x spouse's age	-0.0421	0.0408	-0.0384	-1.0300	0.3030
Enployed p/t	8.3257	1.0386	0.0877	8.0160	0.0000
Sole parent (1=yes)	17.2221	1.0264	0.1912	16.7800	0.0000
Reference person aged 60 years or more (1=yes)	19.0881	0.8537	0.3078	22.3580	0.0000
Spouse employed f/t(1=yes)	-1.7241	1.8558	-0.0243	-0.9290	0.3529
Spouse employed p/t(1=yes)	0.8160	2.0461	0.0117	0.3990	0.6901
Spouse unemployed (1=yes)	2.4222	2.3834	0.0156	1.0160	0.3095
Spouse NILF	4.3167	2.1059	0.0739	2.0500	0.0404
Saturday	20.3509	0.7855	0.2729	25.9080	0.0000
Sunday	24.3030	0.8163	0.3139	29.7710	0.0000
Youngest child 10-14 years (1=yes)	-4.9767	1.1107	-0.0511	-4.4810	0.0000
Youngest child less than 2 years (1=yes)	-7.0562	1.2886	-0.0711	-5.4760	0.0000
Youngest child 2-4 years (1=yes)	-6.7664	1.3421	-0.0623	-5.0420	0.0000
Youngest child 5-9 years (1=yes)	-4.5771	1.1741	-0.0462	-3.8990	0.0001
Equivaqlent disposable income ($)	-0.0006	0.0006	-0.0131	-1.0870	0.2769
(Constant)	30.1194	1.0397		28.9690	0.0000
Adjusted R-squared	0.3103