Family finances

CHAPTER 5

FAMILY INCOMES

The most original, substantial and interesting contribution to the Social Security Review is Alan Jordan's two-volume analysis of incomes in Australia, entitled The Common Treasury:  the Distribution of Income to Families and Households.  Jordan set out to apply the techniques of computer-based cluster analysis to the data from the 1982 ABS Income and Housing Survey, covering 15,000 households and 21,000 income units, with a view to illuminating some very old questions about human well-being.  He lists the questions he has in view as follows:

How are differences in the incomes of individuals and families to be measured?  Why do they matter?  Are we concerned with income as an end in itself, as a means to other ends, as one good among many?  How is income related to other things people value?  Which individuals and families have higher and lower incomes, and why?  How do their incomes vary over time?  Do the inequalities correspond to discernible structures and processes?  When are inequalities to be considered both unnecessary and unacceptable?  Are we concerned only about poverty, or with inequality in general?  In so far as we are concerned with poverty, how do we know whether an individual or family is poor?  If existing inequalities appear to be unjust, what can be done about them? ⁽¹⁾

Economists and other social theorists have attempted to deal with these questions for a very long time, and many have been prepared to accept defeat in the face of them.  Casting their loosely-woven conceptual nets at random into a sea of plankton-sized data, it is not surprising that the nets returned to the boat with little more than the occasional red herring.

Jordan's work is now a useful starting point for discussion of these topics.  His results are significant not just because he has successfully dredged through the data;  he has at the same time raised the standard of interpretation of the subject to a level commensurate with the complexity of the questions.  Previous income analyses have been bedevilled by the fact that the researcher has had to select a priori the criteria deemed likely to reveal significant results.  The Henderson Poverty Commission, for instance, attempted to discover levels of poverty in a select range of "disability groups" -- "fatherless families", "the unemployed", "aged couples", etc.  Jordan's cluster analysis makes it possible for the first time "to sort individuals, income units or households into groups having as much similarity as possible within and as much dissimilarity as possible between groups" with a minimum of rigid presupposition about which categories will and will not produce the most revealing results. ⁽²⁾

The other difficulty which has bedevilled this sort of enterprise is to find some way of comparing people in families with individuals outside the family.  As Jordan observes, "Many more people depend on than receive incomes, and the actual standard of living of the recipient depends on how much goes to the support of dependants, that of dependants on how much benefit is passed on to them". ⁽³⁾  To come to terms with the asymmetry between earners and dependants it is necessary to introduce the concept of "equivalent income".  This notion attempts to provide an approximate measure of standards of living.  To do this it must supply a formula which reconciles three conflicting tendencies:  to allow for the degree to which an income is shared between dependants;  to allow for the different needs of adults and children;  and to allow for the economies of scale enjoyed by those who share a household.  Equivalence scales are calculated not simply on the actual expenses of living, which will vary greatly between rich and poor families, but on the costs of basic needs such as food and housing.  They attempt to measure some minimum standard of living.  For this purpose the needs of wealthy children are regarded as the same as those of poor families.

Jordan adopts the square root of family size as his index.  As he puts it, "The simplest statistic having the desired characteristics is the square root of size of the unit and the simplest method of calculating equivalent income is therefore to divide unit income by square root of unit size". ⁽⁴⁾  On this reckoning a family of four requires about twice the income of a single person to sustain the same standard of living, and a family of nine three times the same income.  Jordan's formula has, quite by happenstance, the advantage of being a very convenient mathematical simplification, but other measures might have been adopted, and some of them may be superior.  The ABS equivalence scales, for instance, rank the first child as approximately 40% of an adult, second as 20%, third as 15%, fourth as 10%. ⁽⁵⁾  We shall adopt Jordan's formula in the remainder of this book, not because it is demonstrably better than any other index, but because it seems to be as good as any other.  One might reasonably suspect that the economies of scale alleged by both Jordan's and the ABS formula are somewhat exaggerated, and thus biased against families -- the last child in a family of nine might well feel hardly done by on these calculations.  But the general reliability of these scales is a complex question and it is convenient to take Jordan as our authority on the matter, allowing that future research may come up with a better formula.  (We can note in passing that the cost of child care -- opportunity cost for the parent who stays at home, actual cost for the family that pays for child care -- is not counted in either the ABS scales or Jordan's scheme.)

Jordan's Common Treasury is as notable for what it does not prove as much as for what it does.  Although the questions being asked are ambitious, he is admirably cautious about the limitations of his techniques.  The data available for cross-indexing against (pre-tax) income include age, family status, educational qualifications, occupation, accommodation, and dependency on social security.  Jordan makes it very clear that such research cannot take direct account of personal differences such as motivation, ambition, intelligence, talent or taste.  Nor can it include family characteristics such as inherited wealth, intra-family transfers and quality of family relationships.  Even more overtly social phenomena such as quality of schooling or income minimisation for tax avoidance purposes also slip through the net.  And Jordan freely allows that income and welfare are by no means synonymous.  As he observes, "Inequalities of income do matter, but ultimately only because of their relationship to the distribution of social goods more important than money". ⁽⁶⁾  Within such limitations Jordan's conclusions are of necessity modest ones.  Indeed, he can say that

Ultimately, only partial answers can be supplied to the question of why some people are better off materially than others.  Describing and explaining the distribution of the sum of material and immaterial benefit is impossible. ⁽⁷⁾

In the end, Jordan's "partial answers" revolve around three concepts:  social class, stage in the life cycle and age cohort.  The age cohort effect -- whether one grew up in the Depression or the post-war boom or the 1970s economic slowdown -- is the least predictive of the three, and need not enter into this discussion.  But Jordan's conclusions about social class do bear repetition here, as it is commonly assumed by people concerned with social welfare issues that Australia is in some significant sense a class society.  The notion of class is notoriously elastic, and any discussion of the subject must be imprecise.  Jordan does not deny that there are class differences in this country, but he emphasises the permeability of those differences.  The differences are differences of median income, but they do not appear to be strongly underpinned by other differences.  As he puts it,

We have demonstrated no sharply-defined strata within which people lead different lives.  None exist in Australian society.  Whatever the imprecision of our groups of higher and lower status, one characteristic that may be accepted as an accurate representation of reality is that their incomes overlap. ⁽⁸⁾

Even among the highest income groups, Jordan found, surprisingly, a significant number of persons without formal educational or occupational qualifications.  When classifying income-earners by occupation he found that the 9 per cent of earners with tertiary degrees had an equivalent income of 388;  the 15 per cent of semi-professionals 286;  the 23 per cent with trade qualifications 247;  and the 53 per cent with none 232. ⁽⁹⁾

Even at the bottom end of the social spectrum, Jordan concludes, the working poor "with lowest earned incomes can no more be seen as a permanent and homogeneous underclass than the people at the other extreme can be seen as a superior caste". ⁽¹⁰⁾  The difference between the gross, pre-tax equivalent incomes of the top 20 per cent of income-earners and the bottom 20 per cent is a ratio of three to one.  After tax the difference between the two would be further reduced.  And even those dependent on welfare transfers do not constitute a clearly-demarcated social stratum.

As compared with the working population, non-aged recipients of government benefits tend to have characteristics generally associated with low income -- relatively low educational status and rates of home ownership, for example -- but not in such a degree as to set them apart from people whose incomes are much higher.  That is, they are an underclass in terms of income but not, as far as one can tell, in terms of other characteristics. ... ⁽¹¹⁾

In other words, the unemployed male is likely to be almost indistinguishable from his employed next-door-neighbour, and the female sole parent almost indistinguishable from her married neighbour.  This suggests that what makes a person a welfare recipient is to be found either in those personal and family characteristics which slip through Jordan's net, or else is a matter of chance.  The implications of this for welfare policy can be postponed until we come to discuss sole parenthood directly.  All that needs to be noted here is that assistance to the welfare population does not seem to require wholesale social change of the kind that (for instance) Marxists would like to see.

It might be insisted here by those who favour class analyses that wealth and not income is the appropriate measure of class differences.  Wealth does not fall within Jordan's purview.  Attempts to estimate wealth differentials are sometimes simplistic.  As with income, wealth comparisons need to be corrected for life-cycle effects.  John Nurick has shown that in a perfectly egalitarian society with no inherited wealth and with everyone receiving the same weekly income from age 17 until death at age 75, if everyone saves 5 per cent of their income at 5 per cent compound interest then the poorest 10 per cent of the population -- young adults -- will have about 0.4 per cent of the wealth and the richest 10 per cent -- the elderly -- will have nearly 30 per cent of the wealth. ⁽¹²⁾  He also points out that calculations of wealth may omit items which should be counted:  the family home, superannuation entitlements, life insurance, the contingent value of social security benefits, the value of one's publicly paid-for education, shares in government assets such as defence, roads, hospitals, schools, and publicly-owned enterprises.  Only once all of these complexities have been taken into account can any conclusions be drawn from wealth distributions about the structure of society.  In the meantime we are confined to discussions of income.

But if Jordan's de-emphasis of class differences is unexpected, his emphasis on life-cycle differences is equally surprising.  His three-to-one ratio of equivalent income for top and bottom quintiles of the working population cannot be taken at face value as a direct indication of class differences because it is affected by life-cycle differences.  The bottom twenty per cent are likely to be mostly young working-class families, and the top twenty per cent middle-aged middle-class couples whose children have left home or are living at home and in well-paid employment.

To compare classes it is necessary to compare like with like.  When Jordan analysed married-couple income units with children he found the population tended to divide into two groups, presumably reflecting class differences;  but the interesting discovery here is that the median equivalent incomes of the two are in the ratio of 3:4 -- not a difference that can be held up as an example of a class society. ⁽¹³⁾  Some other categories of the population produced sharper income contrasts, others produced less demarcation;  the 3:4 ratio seems a reasonable summation of the whole picture.  Yet, when he examined life-cycle differences, he found that a very similar order of magnitude holds here.  For instance the equivalent income ratio of all working families with dependent children to all working individuals including those with families is 4:5. ⁽¹⁴⁾  But this underestimates the difference:  to be more exact here it would be necessary to exclude families with children from the second population and to hold age and class (as measured by education and occupation) differences steady.  Jordan's methods did not permit such a disaggregation because, as he remarks, "Single adults are not a definable stratum of the population". ⁽¹⁵⁾  However, if we can assume that family formation and family size are class-neutral, then some rough comparison might be drawn between the equivalent incomes of those who do and those who do not have children.  The ratio turns out to be 3:5. ⁽¹⁶⁾

Of course the population of persons without dependent children is a heterogeneous one, ranging across young adults before they begin a family, older couples whose children have left home and the stratum of permanently childless.  Families with children are located in the life cycle between the first two of these groupings, and it might be guessed that their average gross incomes (not equivalent incomes) would be about equal to the weighted average of the gross incomes of those two groups.  This is a complicated comparison.  Earnings for primary earners are usually lowest at the beginning of a person's career, and tend to level off at about age 45.  For women, most of whom interrupt their careers to have children, earnings are much more variable, but are on average lowest during the childbearing and child-rearing years.  And incomes fall dramatically for both men and women at retirement age, though this is to some indeterminate extent cushioned by accumulated wealth and assets.  The 3:5 ratio, then, cannot be taken as anything more than the roughest of guides to family standards of living.  The reality may be higher, or it may be lower.

Further consideration of these figures suggests that there will be some even sharper distinctions to be drawn.  Childless couples, for instance, can be expected to enjoy very high equivalent incomes, not just because they will both enjoy uninterrupted careers, nor simply because they do not share their incomes with dependants;  they also have the advantage of the economies of scale that come from shared living.  It might be expected that they would typically enjoy a standard of living at least twice as high as that of families with children.  At the opposite end of the scale, single income families will enjoy equivalent incomes considerably lower than the median for all families with children, though again Jordan's analysis does not permit any quantification of this difference.  Contrasting these two opposite ends of the spectrum -- the double income couple without children and the single income family -- it is easy to arrive at an equivalent income ratio of about three-to-one, the same ratio Jordan found between the top and bottom equivalent income quintiles for working income units.  In other words, the decision to have children and to have one partner caring for them full-time at home is a decision to lower one's standard of living to about a third what it could otherwise be.

The main general conclusion of this analysis is that families with children seem to operate on an average standard of living about three-fifths that of the rest of the working community.  As Jordan observes, "those on whom the community depends for its continued existence -- its children -- are allocated low equivalent incomes, at birth perhaps the lowest of their lives. ...". ⁽¹⁷⁾  Or to put the same point differently, those on whom the community depends for the continuation of the species -- its parents -- have low equivalent incomes, at the birth of their children perhaps the lowest of their lives.  Raising a family is a costly business.

Jordan's description can be supplemented by the results of Beggs and Chapman's analysis of The Forgone Earnings From Child-Rearing in Australia, which calculated that "Relative to the lifetime earnings of childless women, one, two and three children reduce earnings by about 53, 61 and 67 per cent respectively".  On average women forgo about $400,000 in lifetime earnings in order to care for children.  If this hypothetical woman had been able to invest these "earnings" at five per cent per annum interest, the lifetime sum forgone would be closer to a million dollars. ⁽¹⁸⁾  These sums, large though they are, by no means measure all of the time and money that parents put in to their work as parents.  One calculation estimates the forgone earnings costs to be 70 per cent of total child-rearing costs. ⁽¹⁹⁾

From all this it seems clear -- even when the necessary cautions and qualifications are added in -- that life-cycle differences are on the whole just as marked as differences of class in Australian society.  Of course, anyone who denies that justice, prudence or charity require any form of redistribution from rich to poor may also wish to deny that government ought to adjust gross incomes so that people on the same earned gross incomes but with different family commitments enjoy similar living standards.  However, the considerations which support the former may not be the same as those which support the latter.  Indeed, it may be easier to accept the latter.  Equalising the living standards of people on similar incomes seems less likely to set up adverse incentives, and it is perhaps more compatible with rewards for ability and effort.

FAMILIES AND GOVERNMENT

Jordan's Common Treasury is an important addition to our knowledge of incomes, and particularly of family finances.  But it deliberately excludes consideration of the impact of government and taxation on income.  It is commonly assumed that government does redistribute income to families.  The most visible mechanisms for doing this are through the tax and welfare systems, and most discussions of the subject proceed to list as examples of Australian family assistance the Family Allowance, the Family Allowance Supplement for low-income working families, the Dependent Spouse Rebate, the Sole Parent Rebate, the Sole Parents Benefit, and the additional payments for children of other kinds of pensioner or beneficiary.

Discussions along these lines are, however, of very little interest in themselves.  The Family Allowance, for instance, constitutes about $1350 million in the Commonwealth Budget, an impressive-looking sum until it is translated into the dollars and cents of the family housekeeping budget.  As Peter Whiteford pointed out in 1986, "In effect, the family allowance program provides no more than a payment of between 25 cents and 50 cents for each meal that a child has in a week". ⁽²⁰⁾  In fact it is much less than this:  about half of the allowance is already paid for by the average family itself in its taxes, and along the way some of the money is lost in administration costs (not to mention other deadweight losses caused by disincentive effects).  Probably most children get more in pocket money from their parents than the family gets from government in its "allowance".

To appreciate the total impact of government on families we must look beyond personal income taxation and social security spending.  What is needed is a full-scale analysis of the interaction between government and different kinds of income unit, and, fortunately, the Australian Bureau of Statistics fiscal incidence study gives us, for the first time, a clearer picture of this interaction.  The results are just as surprising as those from Jordan's research.  We now know that, taking into account all government services and benefits and deducting all direct and indirect taxes, there is no net transfer to families with dependent children.  In other words, everything that families get from government -- including the most costly item, education -- they pay for themselves.  Indeed, they pay for slightly more than they receive.  The only transfer to poor families comes from better-off families.  There is no transfer to families with children from those who never have children.  And for those who do have children there is no transfer from the periods before and after they have children to that period in the middle when their needs are greatest.

The Bureau's findings on the interaction between government and families are summarised in Figures 5.1 and 5.2, which rely on a paper by the Australian Statistician, Ian Castles. ⁽²¹⁾  Castles selects ten broad, mutually-exclusive household types from the survey population, defining them in terms of family structure and age of household members (explanations beneath Figure 5.1).  The charts show the average benefits received and taxes paid per household for each of the household types, and the net effect.  Figure 5.1 shows dollars per week (in 1984), and Figure 5.2 shows the amounts as percentages of final income.  The width of each column in the figures is proportional to the number of households in each category (though not to the numbers of persons in each household type).

Figure 5.1:  Benefits & Taxes per Household
for selected household types (average, $ per week)

Composition of the selected Household Types:

1. Young single (one person living alone, age under 35).
2. Young married couple (no children, household head aged under 35).
3. Couple and child(ren) under 5 years.
4. Couple and child(ren), eldest aged between 5 and 15).
5. Couple and dependent child(ren), eldest aged 15 or more.
6. Couple and both dependent and non-dependent children (but no other household members).
7. Couple and non-dependent child(ren).
8. Middle-aged couple (head aged between 55 and 64).
9. Elderly couple (head aged 65 or more).
10. Elderly single (one person living alone, aged 65 or more).

Source:  Table A in I. Castles, Government Welfare Outlays:  Who Benefits?  Who Pays?, Canberra Bulletin of Public Administration, March 1987.

Figure 5.2:  Benefits & Taxes per Household
as per cent of final income, for selected household types

See Figure 5.1 for information on household composition.

Source:  See Figure 5.1.

The Bureau found that the total revenue taken from direct and indirect taxes roughly equalled total government "social" expenditure -- that is, outlays on such items as education, health, housing and social security, each of which is substantially a private good.  Conveniently for our purposes, the remaining expenditure -- the "public good" budget -- can be thought of as coming out of other forms of taxation.  Taken overall the graph indicates that the main net transfer in the Australian social system is from young to old.  Pensions and health services for the aged are paid for mainly by income earners below the age of 35, some of whom have young children.

Our particular interest is in households with dependent children -- types (3), (4), (5) and (6) -- amounting to about half the selected households.  As the shaded net effect areas show, these groups taken together give more to governments than they get back, and the major contributors are families with young children.  The only net beneficiaries are type (5), couples with older children of secondary school or tertiary education age.  The charts enable us to modify Jordan's broadly-correct observation that "society is not much concerned with young children, shows more interest when they go to school, and invests most heavily in those who, the time approaching for them to become independent agents, disclose plausible claims to higher status". ⁽²²⁾  Society does indeed allocate its resources in this way, but the resources are those it has taken directly from the families themselves.  Education is the only major item of government spending which goes primarily to families, and both primary and secondary education are paid for entirely out of family taxes.

The ABS fiscal incidence figures can be supplemented to some extent with information about changes in personal income tax payments, as collated in a recent EPAC paper, Income Support Policies, Taxation and Incentives.  No straightforward connections can be drawn between the two sorts of data because income tax figures tell us nothing about government spending.  One social group may have its tax payments increased and yet be a net gainer if government spending is directed towards it and away from other sectors.  And vice versa:  a group may enjoy tax reductions and yet still lose out if government spending is directed away from it.

In 1954-55 a single taxpayer (someone with no dependants) earning average weekly earnings paid 9.60 per cent personal income tax;  an earner with a spouse and two children to support paid 4.96 per cent, but after child endowment payments are taken into consideration this amounted to only 0.57 per cent.  In 1964-65 the same comparison gives figures of 14.36 per cent, 9.32 per cent and 6.57 per cent.  In 1974-75 the figures have jumped to 21.96 per cent, 17.30 per cent and 16.29 per cent.  Ten years later the figures are 24.59 per cent, 19.96 per cent and (we are now counting Family Allowance, not child endowment) 16.97 per cent. ⁽²³⁾  Another way of putting this is to say that after family assistance is taken into account, the proportion of tax paid by average-income families to that paid by average-income single persons is 6 per cent in 1954-55, 46 per cent in 1964-65, 75 per cent in 1974-75 and 69 per cent in 1984-85.

Between 1964-65 and 1984-85 the net tax (counting family assistance) paid by average single persons increased by 71 per cent and that by average (two child) families by 158 per cent.  If we compare individuals and families on 0.75 average earnings then the contrast is even more dramatic:  78 per cent and 325 per cent.  At 1.50 times AWE (61 per cent and 102 per cent) and at 2.0 times AWE (58 per cent and 84 per cent) the comparison is somewhat less stark, but the same trend is clear.  While the 1989 May Economic Statement, which increased and indexed Family Allowances, will have made some slight difference to these figures, the difference does not significantly alter the long-term trend.

The taxation picture can be further sharpened if we distinguish between two-parent and one-parent families.  An earlier EPAC publication, Aspects of the Social Wage, demonstrates that whereas two-parent families are net contributors to government (thus confirming the ABS finding), sole parent families receive in government benefits four times as much as they contribute in taxes. ⁽²⁴⁾  Since one segment of the population of families with children is so strongly favoured by government assistance, and as families on the whole get no net benefits from government, we can conclude that two-parent families generally lose from their interaction with the tax/transfer system.

International comparisons are sometimes employed to support the conclusion that Australia gives little support to families with dependent children, but this evidence is misleading.  Carol Oxley surveyed the international family assistance scene for the Social Security Review.  Her contribution, The Structure of General Family Provision In Australia and Overseas:  A Comparative Study, provides evidence for the conclusion that (contrary to common belief, and contrary to Oxley's conclusion) Australia appears to be, by international standards, not ungenerous to families with children.  The belief that Australia is ungenerous rests on a comparison between levels of universal family assistance in the OECD countries. ⁽²⁵⁾  On this comparison Australia ranks twelfth in a list of sixteen countries, providing about half the international average level.  A more revealing comparison, also based on 1984 figures, is that between the levels of family disposable income after tax. ⁽²⁶⁾  On this way of reckoning the matter Australia ranks sixth in a list of thirteen, considerably higher than Sweden and the Netherlands, who far outrank us in the previous table, and not far behind Belgium and Austria, the top two in the first table.  Even more curiously, Spain and Canada, who were last and thirteenth in the first table, come out fifth and third in the second table.

Oxley does not discuss these remarkable transpositions.  One obvious interpretation is that those countries who rank highly in the first comparison but low in the second, such as Sweden and the Netherlands, give generously to families but then tax back almost as much as they give, a perfectly pointless exercise known to policy analysts and economists as "churning".  On this comparison Australia's low level of family assistance at least can be credited with being well designed, relative to international standards.  However, the best test of family assistance levels can come only from full fiscal incidence comparisons, and (as we have seen) Australian governments provide no net assistance to families.  If this fact is reflected in Australia's ranking in Oxley's second table then it would appear that many advanced countries are even less helpful towards families than we are.  But, as Oxley observes, these figures take no account of whether or not government spending favours families.  In general it seems best to conclude that the international comparisons presently available are difficult to interpret and of little value for policy purposes.

What is clear is that the extraordinary growth of government in Australia in the past few decades has been financed disproportionately out of family earnings, particularly the earnings of two-parent families on PAYE taxation.  Much of this growth has taken place in the welfare sector, if we interpret that term broadly, to cover health, housing and education as well as income transfers and welfare services.  The paradoxical conclusion suggested by the above evidence is that the expansion of the welfare state in the past two or three decades has been paid for mainly by the primary welfare institution, families with dependent children;  but in return for their contribution they have received very little, even after taking into account the rapid increase in expenditure on education.  In other words, one welfare system -- the state -- has tended to displace or impoverish another -- the family.  If, however, we attribute the growth of government to a more general process of capture by interest groups or (in Olson's terms) "distributional coalitions", of which the welfare industry is only one amongst many, then the evidence reinforces the argument of Chapter One that families are poorly equipped to compete in this sort of politicking.  This trend away from universal family assistance runs counter to the general trend which Goodin and LeGrand have described as "creeping universalism in the Australian welfare state", as evidenced by trends in the age pension and the invalid pension. ⁽²⁷⁾  A further generalisation suggested is that, while Australia is not a class society, it may have an unbalanced allocation of resources between the generations, making the child-rearing years even more difficult than they intrinsically are.  This point will be pursued in the later part of this chapter.  Chapter Six will consider some of the further implications for taxation policy.

COPING STRATEGIES

So far, then, we have found -- with as much assurance as any such matter can be known -- that families operate on about three-fifths the equivalent income of the rest of the community;  that they get no net advantage from government;  and that the personal taxation system has been steadily eroding whatever advantages families once enjoyed.  At this point, however, opinions diverge even amongst those who will accept this description of the situation.  Some will come to the conclusion that this is a society which wants to make parenthood as economically unpleasant and unattractive as possible.  Others mount a case for saying that this is how the social system should be arranged.  But before we enter this debate it is necessary to note the way in which, in the absence of any social support, families have responded to the economic difficulties they face.

There are four obvious strategies that couples can adopt to cope with these economic pressures.  One is not to have children at all;  another is to limit the number of children they have;  a third is to delay having children until an adequate capital base has been built up;  and the fourth is to take a second job while caring for children.  It is the last of these which has occasioned most discussion, though all four have become evident social trends at least since the early 1970s when economic growth began to slow down.  While it cannot be assumed that these trends are caused solely by economic factors, there is no reason to doubt that economic factors play some part in bringing them about.

The first two of these strategies is indicative of a world-wide trend amongst industrialised societies towards population deceleration or decline.  The full explanation of this trend is far from clear, and its interpretation from a policy viewpoint is also the subject of discussion in some countries, though not yet in Australia.  Whether social policy should -- or can -- play some part in preventing population decline is a question which might become more urgent if the present trend continues.  A book like this is not the proper place to enter into this debate.  The only assumption which will be made here is that Australia, which presently has an active immigration program, can at least seek to sustain something close to its present population level.  In other words, a policy which supports families need not be feared on the grounds that it will contribute to the depletion of resources.  Family support policies are known from European experience to have rather little effect on these trends.

The third family strategy is postponement of parenthood.  The trend has been for teenage birth rates to fall slightly but steadily, for birth rates for women in their twenties to remain stable, and for the birth rate of women in their thirties to rise steadily but slowly.  Over the whole life span of any particular family there is presumably no net economic gain from this postponement, for the early completion of the child-rearing phase brings compensating gains to those who have their children when they are young.  Postponement does have the advantage of making the child-rearing years economically easier, though it is, even so, an advantage of limited duration.

Women in the Home (the Victorian Women's Consultative Council's 1986 report mentioned in Chapter Four) found, as we saw, that the most pressing problems of homemakers are lack of money, especially lack of personal income, lack of status and lack of leisure.  Money and status correspond to the difficulties of family economics and the decline of family morale which are central themes of this book.  Complaints about lack of money should not necessarily be taken at face value -- experience sometimes suggests that the more money people have the more they complain about the lack of it.  But if, as we saw, the average family operates on an equivalent income three-fifths that of the average person without dependants it seems plausible that this reported "lack of money" is not merely the self-interested or self-indulgent grumbling which most of us engage in from time to time.

It might seem that the reported lack of personal income is best solved by encouraging women to enter the workforce, and there is no doubt that many women who do take a job appreciate the financial independence that a second income brings.  However, although "lack of personal income" appeared as the most frequent complaint overall and "family poverty" as only the fifth most frequent complaint, when we examine the ranking of "main difficulties" "family poverty" rates almost as highly as "lack of personal income" (still the highest).  There is, it seems, no clear distinction between personal and family hardship.  If the somewhat artificial distinction between the two sorts of need is removed then the combined category ranks far above all others.

Furthermore, and surprisingly, a second job was not listed by these homemakers as a priority.  This may be a consequence of the design of the questionnaire, which did not specifically ask respondents about aspirations to employment;  it did, however, leave space for open-ended comments, and some of those comments indicated some concern about loss of self-confidence about returning to the workforce.  However, although we have little reliable research on this subject, other studies tend to reinforce the conclusion that many mothers who are working would prefer to spend more time caring for their family.  The 1984 Clemenger/Reark study of working women (not specifically mothers) concluded that 64 per cent worked to earn an income and 36 per cent for career reasons. ⁽²⁸⁾

The extent of women's entry into the paid workforce is often misconstrued.  Discussions of this subject need to distinguish between part-time and full-time work, and between workforce participation by all females, by married women, and by mothers of dependent children.  The most commonly mentioned statistic is the fifty-something per cent of married women in paid employment, a figure that has risen steadily in the past three decades.  This figure may disguise as much as it reveals.  What is less well known is that only about 12 per cent of mothers with children under five years of age work full-time, and only about 25 per cent work part-time.  Less than 50 per cent of all mothers of dependent children are in the workforce, only 20 per cent of them full-time and 27 per cent part-time.  A few per cent are classified by ABS as "discouraged job seekers". ⁽²⁹⁾

As a general policy prescription for family needs, "Let them get a job" has all the subtlety and appropriateness that Marie Antoinette's "Let them eat cake" had for an earlier generation of housewives.  To ignore the wishes of homemakers in this matter is to continue to belittle their status in the manner deplored by the women respondents to Women in the Home.  If they do adopt this solution they may merely substitute emotional and physical stress for economic hardship.  Even the financial gains of this strategy are often rather limited after childcare costs are taken into account.  Many mothers work more from necessity than preference, and to add another job to the family workload does not make the primary job go away, however it may be shared between husband and wife.  To say this is in no way to denigrate mothers who make the decision to work primarily for their own needs and interests;  it is only to mark a distinction which is essential to understanding the present situation of the majority of families.  Rather than offering a generalisable solution to family financial difficulties, entry into the workforce is often like stepping onto a treadmill where, as in Alice's Wonderland, it takes all the running you can do to stay in the same place.

In considering this topic Jordan observes that, although "the employment or non-employment of married women is a very important determinant of the relative incomes of families, [it] has not been and is not an unambiguous indicator of welfare.  It can indicate poverty. ..."  He goes on to add that

The working woman's net contribution cannot be assessed, and neither can the net effect of employment on her own level of welfare when, as seems to be true of every country in which female labour-force participation has risen in the last generation or so, domestic demands have not declined in proportion to external activity.  Probably, that is to say, the visible economic gains have often been offset by a loss of leisure. ⁽³⁰⁾

Writers on this subject often assume (though Jordan may not) that the alternatives here are work and leisure, but a woman who takes a job to supplement the family income may be losing other things even more valuable to her and her family, such as time with her children.  Whether or not this time is to be deemed "leisure" is a question of linguistic nicety, but the substance of the distinction is a very much more important matter.

FAMILY ASSISTANCE

It seems then that there are no easy ways for families to offset the costs of children which can be implemented by families themselves.  But as we have seen the disparity between the "needs" of individuals and those of families is not -- or not any longer -- given any special recognition in wages, in government expenditure, in the taxation system, or in the welfare system, except in the form of additional assistance to pensioners and beneficiaries with children and in the Family Allowance Supplement for low-income working families.  All other government concessions to families or expenditure on families are likely to be wholly paid for out of the taxes of the recipients.

Although this belief is no longer embodied in our taxation and social welfare system, most people assume that family assistance is a justifiable form of social expenditure.  An AGB McNair survey for the Australian Institute of Family Studies, for instance, found that eighty per cent of respondents agreed that families with children should get more tax relief than families and individuals without children. ⁽³¹⁾  The explanation of the discrepancy between common belief and current practice is perhaps threefold:  the disadvantages suffered by families (as detailed above) have not been made widely known;  the political process (as was argued in Chapter One) has been commandeered by lobby groups with interests indifferent to family needs;  and (as was suggested in Chapter Four) the "New Class" elite of policy-makers and opinion-formers has often been unfriendly to family aspirations and achievements.  In such a situation what was once taken for granted must be defended.

One way to approach this issue is to argue that the conventional criteria of good tax design, in particular the commitment to horizontal equity, require a system which recognises family needs.  This is the approach taken by Jordan, and it will be pursued further in the next chapter, but before taking up that theme it is necessary to look at the question of family support in a broader perspective.  In the larger framework there are two questions which must be answered:  Is it fair in welfare and tax policy to recognise the costs which are peculiar to families?  And is it prudent to do so?  Realistically the second question is the more important of the two.  The suggestion that those who do not have children might support those who do, however persuasive, is not likely to raise much revenue by itself.

Family support can be viewed as a life-cycle transfer.  This amounts to a "childhood pension" something like the age pension.  The taxpayer effectively uses the tax and welfare systems to equalise his or her standard of living across the life-cycle.  The proportion of permanently childless in the population is small, amounting to about 15 to 20 per cent of childbearing age adults.  Most of the support for people with dependent children will have to come from those income earners who have not yet begun families of their own and from those earners who have finished raising their own family.  But these are the very people who will benefit in future or who have benefited in the past from family support schemes.  As a matter of prudence the balance of gains and losses from such a scheme needs to be calculated rather carefully.  Economists argue that all income transfers entail both macro- and micro-economic losses;  we need to make some estimate of the net benefits that might result from a proposal such as this.

Before considering the question of economic costs we need to lay out the advantages of the proposal.  This is a matter we might have expected the Social Security Review to have discussed very thoroughly, yet it nowhere attempted to weigh up the merits of any substantial system of family assistance.  An early Discussion Paper by Ann Harding announced that "The question of the appropriate balance between universal and income-tested family assistance programs ... will receive detailed consideration during the Review", but this promise was never carried through.  A subsequent paper by Peter Whiteford specifically on Issues in Assistance for Families -- Horizontal and Vertical Considerations made no attempt to consider the case for a very much more substantial system of horizontal equity assistance than that presently in existence. ⁽³²⁾  Considering the resources available to the Review, this general timidity seems quite unaccountable.

The obvious rationale for family support has to do with the nature of the human life cycle.  For most people becoming a parent entails a financial crisis brought about by four factors:  children themselves involve special needs and recurrent costs;  at the same stage the family's "capital" costs of housing are at their most acute;  with the birth of children it becomes difficult, and in many cases not desirable or desired, for both parents to be in the paid workforce;  and at this time the main breadwinner is usually at the beginning of his or her career, and is thus not likely to be earning a high income.  The problem is essentially biological:  if parenthood was normally entered in one's forties or fifties, no such crisis would arise.  The financial crisis is, however, only a cash-flow problem.  Over the whole life cycle almost all families are capable of being financially independent.  A system of family support would even out the cash-flow difficulties.  It would benefit us all when we are children;  and it would benefit parents of dependent children when their income is low relative to their needs.  It thus makes families not less but more self-supporting.

Family assistance is both a subsidy from later life to the child-rearing years and a form of compulsory saving before becoming a parent, and both of these aspect require consideration.  It is an important fact that most parents earn a higher income in later life than they do in early adulthood, especially if in in later life both are in full-time employment.  The only other way to achieve this kind of subsidy would be as a loan on future earnings, and, since few young adults would have the resources to back a loan of this kind, it is unlikely that a bank would accept such an arrangement.

Discussion of the compulsory saving aspect should keep in mind the small fact that age tends to bring maturity and wisdom.  Many parents no doubt regret the money that slipped through their fingers when they were young.  The ABS Household Expenditure Survey found that adults under thirty save only 0.7 per cent of their income, the lowest savings level of all age groups apart from the 55-65 age group (many of whom are using lump-sum superannuation benefits to finance early retirement).  Jordan found that young adults before becoming parents enjoyed relatively high equivalent incomes. ⁽³³⁾  Some of this group will be using their savings to buy a house, but this does not explain how it is that the 30-44 age group, most of whom also have dependent children and high housing costs, have a much higher savings rate of 3.4 per cent.  Paternalism perhaps has a place here.

Libertarian opposition to family assistance entails also the elimination of age pensions, which are themselves a form of compulsory saving.  It is not impossible that in a purely libertarian society, a society which abolished the whole creaking, rickety Heath Robinson apparatus of state-funded welfare (from age pensions and tax-supported superannuation down to discounts on bus rides for school children), the great majority of families would be better off than they presently are in a Welfare State which recognises all sorts of needs except those of families with dependent children.  However, it is not easy to know much about a state of affairs so radically different from that which presently obtains, and there is little political will to abolish all state welfare (a point well made by Lawrence Mead against abolitionists like Charles Murray), so the issue is somewhat academic.

The present trend, particularly amongst the more highly educated, to later marriage and postponement of child-bearing expresses in part a desire to provide well for their future family.  Family assistance would make the accumulation of wealth before having a family both more difficult and less necessary.  (Only those who thought they were saving for a family and then later decided not to have children would lose by this proposal.)  This trend is also in part the result of a desire to get established in a career before facing the "interruptions" of parenthood.  There would be no net cost imposed on any person who took this course, but they would suffer a loss of choice in how their savings could be invested.  The loss of choice in investment is balanced by the gain in choice about when to have one parent at home full-time.  This option would not exist if family assistance had to be privately funded.  One (perhaps minor) benefit of family assistance is that it makes easier the opposite decision, to have children early and pursue a career later in life.  At forty a person still has about twenty working years ahead of them, and the advantages of added maturity.  Whether at present employers can see these advantages is perhaps questionable;  it is reasonable to think that, if they are in their interests, they will in time come to see them.

In the phase-in stage those who have finished raising a family would lose from the scheme.  It would require some sacrifice on their part, perhaps motivated by memories of the stringencies they experienced as parents.  Most of the present older generation enjoyed the low tax era of the fifties and sixties, but they will still remember vividly the general difficulties of parenthood and could see the scheme as a form of assistance for their own grandchildren.  The AGB McNair survey mentioned above found, surprisingly, that people over 45 were as likely to support tax relief for families with children as were the families with children themselves. ⁽³⁴⁾

OBJECTIONS TO FAMILY ASSISTANCE

We can turn now to the economic case against family support.  Before entering into the detail of this issue it is worth remembering that social practices vary widely between societies.  Some societies such as France, without any apparent loss of economic efficiency, employ generous family assistance schemes.  Australia, which does not have a good record for economic efficiency, has a social system which provides very little family assistance.  On the face of this evidence there would seem to be no direct relation between the two.  But we need to look at the issue in greater detail.

At the macroeconomic level there is obviously reason for considerable caution about any large social policy changes.  With a net external debt equivalent to more than $6000 per head of population we need to be sure that nothing is done to exacerbate our precarious economic situation.  As we have seen, most of the growth of government has been financed out of family earnings;  but much also has come from government borrowings.  In effect this has jeopardised the economic future of the present younger generation.  It might seem that, through the imprudence of government in the past two decades, we are now faced with a cruel dilemma.  A policy which has reduced the means of families unjustly and unwisely should but cannot be undone because the cure will be even more ruinous than the disease.  This objection requires that we take seriously both the macro-economic and the microeconomic consequences of any policy changes.  And it suggests that perhaps the best reason for economic reform -- debt reduction, productivity improvements, export campaigns, tariff reductions, wage restraint, labour market deregulation, tougher scrutiny of the welfare system -- is to restore equity to family finances.

The microeconomic objection to family support can be put as follows.  The costs of such support require tax increases, which will raise the effective marginal tax rates (EMTRs) for persons without dependants and perhaps also for families or secondary earners in families.  Higher marginal tax rates lower labour supply, as people come to prefer leisure to work.  The net effect is a loss to both the givers and the receivers, and a loss to the economy as a whole.  The details of this argument are a technical matter, but some general comments can be made about the hidden assumptions at work here.

There are two such assumptions.  One is that primary income earners with dependants to support are more likely to lower their labour supply than individuals without dependants.  Take two people earning the same income, one with five children and the other with none.  The first enjoys a standard of living less than half that of the other, even (under the present regime) after tax.  Their EMTRs are identical (except for those low income taxpayers receiving the Family Allowance Supplement).  Who is more likely to take on some offered overtime?  Very obviously the first.  Much as he (or she) will value the possibility of some leisure time with his family, their very pressing needs will force his decision.  But suppose we make significant welfare transfers to him, so that his net income gives him and his family an equivalent income about two thirds that of the other person.  His EMTR is increased by the income tests (if any) that govern the welfare payments, and is now significantly higher than his colleague's.  Who now will take the overtime?  It is not clear that even now the person with dependants is less likely to do so than the other person.  And of course if the transfers are made in the form of tax concessions then his EMTR will stay as before.  So there does not appear to be any necessary loss of efficiency at this point

But families contain two potential income earners.  Maybe it is the effect on the actual or potential secondary earner which causes the loss.  Here the second assumption comes into play.  She (or he) will be making a choice between working at home and working in the workforce.  Is this choice to be thought of as a choice between work and leisure?  Homemakers with dependant children do not commonly think of their efforts as leisure. ⁽³⁵⁾  It seems then that any change in her EMTR will affect only her choice between one kind of work and another.  There is no reason to assume that one is "real" work and the other not, or that shifts between one and the other in either direction add to or subtract from the net wealth of society.  The fact that only one kind is counted in the GDP is little more than an indication of the limitations of national accounts.

Of course it may be in some particular cases that a mother would be better off in the paid workforce, but economists are in no position to estimate how many such women there are in the community.  And to count such cases fairly we would need to count also those women who are presently working but would be better utilised at home.  Until economists know as much about what goes on in the ordinary family home as they do about the business world they should leave judgements about such decisions to the persons who make them.  If it is the case that some parents would be more usefully employed in the home than in the labour force then there would be some economic gains from family assistance.  Apart from the advantages of improving the quality of life within families it might also lower unemployment by making available jobs presently held by two-earner families.

Other efficiency gains may be achieved by channelling family assistance through the tax system instead of through allowances.  Economic losses caused by "poverty traps" which deter welfare beneficiaries from seeking employment are difficult to calculate, but in a society with some 700,000 workforce-aged unemployed and sole-parent beneficiaries -- nearly a tenth of the labour force -- the advantages of eliminating these disincentives should not be underestimated. ⁽³⁶⁾  This theme will be pursued further in the next chapter.

FAIRNESS

The second question concerns the fairness of family assistance.  Claims about the importance of the family sometimes figure prominently in the rhetoric of politicians, and will probably do so with increasing frequency in the next few years.  "The family is the basic unit of society", we are told, but it is not at all clear how rhetoric of this kind translates into useful and fair social policies.  In particular, it is not easy to see how a system of life-cycle transfers to help families through the child-rearing stage can be implemented without taxing those who never have children to help pay for the system, and it is difficult to see how those taxes are to be justified if they are unavoidable.  At issue here is the question of whether assistance to families with children is anything more than an arbitrary imposition by yet another special interest group upon that section of the public who for whatever reason remain childless.  There seems to be a clash here between the economic interests of children, which might be assisted by government policy, and the economic interests of the childless, which unlike the interests of children are catered for in the market.  Can this clash be mediated and moderated by rational argument?  The question requires us to ask some further difficult questions about the nature of families and about basic principles of public policy.

In our society it is basically families which create the next generation;  without their work there would be none, and no social continuity.  Some families do this job better than others, and a few do it badly, but most try to do it well and the majority succeed in what can be an exceedingly difficult task.  But if this is so then it seems plausible to suppose that those who do not have children owe a debt to those who do.  Social continuity is valuable to everyone who hopes to see the achievements and projects of their lifetime preserved and perpetuated.  Thus, given the assumption that most families do in fact contribute positively to society, one kind of case for family assistance is based on the work which families do.

It is difficult to see how an argument of this kind could be adjudicated, but one way to consider the question is to ask whether there is any realistic alternative to the family.  Some alternatives have been tried, but the evidence suggests that every alternative gravitates back to the old parental model.  The Israeli kibbutz movement, which never severed the connection between parent and child though it did attenuate it, has tended with time and experience to restore the parent-child relationship.  Parents found they wanted to enjoy the specialness of that relationship.  But, more importantly, it was discovered that if the bonds between parents and children are weakened the children form equally powerful but usually less benign attachments -- in their case peer-group attachments.  The kibbutz children, even with enormous social investment in their well-being, did not develop into the autonomous individuals that their society expected to produce.  As Brigitte and Peter Berger have put it,

These [children] are individuals who find it extraordinarily difficult to stand up against their group, who find it difficult to develop an inner life outside the sphere of collective activity, and who very often find it hard to exist in any less collectivist situation (such as the rest of Israeli society, outside the kibbutzim). ⁽³⁷⁾

Other communal experiments with child-rearing have in general been very much less successful than families generally are.  Many of the 1970s experiments were consciously in rebellion against "the family", though they sometimes borrowed its rhetoric, but very few have survived and demonstrated their viability.  Even fewer can be counted as clear successes.  The reasons for this failure are not obscure:  child-rearing is very time-consuming and sometimes tedious, and few adults will persist with it conscientiously unless they enjoy close and reciprocal emotional bonds with the children, of the kind taken for granted in the average family.

The same problem arises in state-supervised substitute care for children who have lost or been rejected by their parents.  In general, the state has found it preferable to relocate such children in other families than to create alternative childcare institutions.  The best alternative institutions may be moderately successful, but they are so by mimicking the family.  They serve as surrogate families, not as alternatives to the family.  And (to keep the extent of state supervision in perspective) only about one child in every three hundred requires state-supervised care, whether through adoption, foster parenting or institutional care;  the other 299 are with their parents or one of their parents.  Of these 15,000 children about one third were in institutional care and one half were with foster parents. ⁽³⁸⁾

The universality of the family makes it and the work it does almost invisible.  We notice it only when it is missing.  If there were other ways of doing that work we would certainly know about them, for the family has had more than its share of enemies.  But there are no alternatives, because the family does the job vastly better than anyone else.  No matter how much we might argue about any number of its many apparent defects, the family seems to continue to hold the field without any serious challenger.  If we want to sustain our social world then, it would seem, we should be prepared to recognise the organisation which preserves the continuity of that world.

We return to the original question:  Should those who do not have children be required to help those who do?  The case for denying any such obligation is usually put as follows:  The decision to have or not have children is just like any other consumer choice.  Those who make the decision or who take the risk of becoming pregnant are fully aware of what they are letting themselves in for, or if they are not they should be.  Children who really add to the general good of society are few and far between:  the majority are just extra members of society, and are on the whole neither desirable nor undesirable additions to the common stock of humanity.  Some are unquestionably a burden on the rest of the population.  Even if the newcomers were on the whole beneficial, their being added to the population entails no obligations upon those who never sought them in the first place.

Any adequate response to this kind of objection will need to engage in the wider debate about the nature and value of the institution of the family, a debate well described in the Bergers' The War Over the Family.  The Bergers do not discuss family finances, but we know that (in Australia at least) finance appears to be a critical difficulty for many families.  But, as the above objections show, the question of financial support for families cannot be considered in isolation from the larger debate.  Assistance schemes which are optional, or which do not involve those who never have children of their own, are difficult to design and achieve.  Any family assistance scheme must, therefore, be acceptable to those who do not benefit directly from it.

The issue here is not simply one of fairness, because many who will not benefit directly may be happy to waive their own claims in favour of promoting what they perceive as conducive to a more harmonious and constructive society.  This is, arguably, what we already do -- or aim to do -- when we accept programs which tend to equalise incomes between rich and poor.  Jordan's analysis of incomes provides strong evidence that at present much of the contrast between rich and poor is a contrast between those with dependent children and those who never have children.  Shifting our welfare system towards a family assistance system may be a way of putting that equalising intention more effectively into practice.  It would also, as we saw, tend to make families more effectively self-supporting over the whole life cycle.

The case for family assistance would be more impressive if there were good evidence that financial aid would improve the social performance of families.  Ideally, a two-fold campaign is needed which both assists families financially and reinforces their status, authority and stability in a manner that tends to reduce the socially-destructive anomie arising from family failure so evident at present.  Part of the case for believing that this can be done is made out in The War Over the Family.  The authors (both distinguished sociologists with a strong interest in child development) contend that the empirical consequences of the pressures upon families today include "juvenile delinquency and (increasingly) serious crime, drugs and alcoholism, suicide, a frenetic preoccupation with sexuality, mental disorders, and the appeal of fanatical cults". ⁽³⁹⁾  Their book is a defence of what they call the "bourgeois family", a unit characterised by two adult parents providing children with the stability, authority and love required for the development of a secure identity, capable of resisting disintegrative forces at work in the larger society, and often enough within the family itself.  ("Bourgeois" here is more a cultural than a sociological term;  the authors believe that the "bourgeois" family is now more likely to be working class than wealthy middle class.)

The Bergers' claims are to some degree sociologically testable.  In Australia, however, research on such matters is either non-existent or somewhat primitive. ⁽⁴⁰⁾  At present the questions have to be considered in the subjunctive.  The case for family assistance rests in part on the Berger hypothesis that such a scheme could play an important part in preventing the further formation of a violent, aimless, self-destructive and costly underclass.  We have good reason to doubt that existing welfare policy is arresting this process, and good reason to believe that it may be promoting it.  Subsidising families who take their social responsibilities seriously may even be a more cost-effective approach than that presently employed.  If it is true that the Bergers' "bourgeois family" produces "responsible and autonomous individuals" more successfully than any other social arrangement, then a society which wanted to produce such individuals would have good reason to assist that kind of family, and the value of this assistance might well be apparent to those who do not have children of their own. ⁽⁴¹⁾  If it were possible to support only the kind of family which produces those individuals then social policy might seek to give selective assistance.  At present a selective approach is indeed being adopted, but it is for the most part selective in favour of a family model which does not generally meet the Bergers' criteria of family success.  It may not be possible or desirable to reverse this selectivity but there is no obvious reason why assistance to the Bergers' "bourgeois family" might not be equalised with that to the level enjoyed by single-parent families and non-custodial parents.

Not all those who would oppose such a family assistance scheme would do so for self-interested, short-sighted or sophistical reasons.  In areas as controversial as this a variety of intelligent viewpoints may be held.  Some may believe quite sincerely that families with children neither need nor deserve social assistance.  However, we cannot know what public opinion will be on this matter until it is tested in the political marketplace.  One reason for thinking that it might be politically acceptable is that it has been commonly assumed that families with dependent children do benefit from government;  the discovery that they do not is recent and not yet widely appreciated.  Increased assistance to families may only put into practice what has been assumed, falsely, to exist already.  But in politics it is impossible to please everyone, and even when the case has been made as well as it can be, there will still be those who, even with good reason, may oppose such a policy.  How significant such opposition is has to be decided by politicians, by public debate and by the electorate.

Of course family assistance is not just assistance to parents.  Those who never have children benefit when they themselves are children.  Another consideration is that we presently subsidise immigration to this country, which suggests that we do set a positive value upon new members of our society, even if not upon new members produced by the present membership.  Further, it may not be necessary to assist families throughout the whole of their children's childhood.  It may be sufficient to provide generous family allowances up until the child turns (say) ten.  Although the recurrent costs of children (food, clothing, etc.) increase with age, the capital costs, particularly housing, are also greatest in the early stage of the child's life.  The constraints on parents' ability to work are greatest during the early childhood years.  Another reason for weighting assistance to the early years is that at present families are assisted to a small degree in the later, secondary school years, whereas in the early pre-school years governments take more than they give.

Whatever the decision on the appropriate age criteria for recipients, any scheme which seeks to have a significant impact must be by present standards very generous -- say something in the order of $50 per week for the first child, $40 for the second, $35 for the third and $30 for all subsequent children.  Such a scheme could be assets-tested just as the age pension is assets-tested, perhaps even on the same financial basis.  The total cost of a scheme restricted to under-ten year olds would be about $6 billion, or about $3 billion net allowing for present expenditure on Family Allowance, Family Allowance Supplement, and other forms of family expenditure.  (That is, roughly equivalent to what we now spend on keeping unemployed people -- or keeping people unemployed.) ⁽⁴²⁾

Family support of this magnitude would represent a very large change in social policy.  The Social Security Review laboured mountainously in this area and, in the end, seems to have brought forth only the squeaking mouse of Family Allowance indexation.  The Review's failure even to debate the relevant considerations should not be taken as proof that there is no case to consider.  What is needed is rational debate about such a proposal by both economists and the public at large.  The argument presented in this chapter should be read more as presenting the case for such a debate rather than as presenting the case for any particular conclusion, but it should at least be evident that there is a case for family assistance which needs to be considered.

Broadly speaking, the proposal being advanced in this chapter for a system of assistance for families with dependent children is similar to the existing system of age pensions, and raises much the same issues of justification and design.  Both systems may entail some economic loss.  That loss needs to be justified in terms of social benefit.  Both must be designed to strike a fair balance between contributions and needs.  One way of justifying such a scheme for children is to compare the standard of living enjoyed by dependent children and their parents under the present regime with that enjoyed by age pensioners.  If it turned out that the elderly were markedly better off than the young then considerations of parity would suggest some action to equalise their situations.  No such careful comparison seems to have been carried out in the Australian welfare literature.  Some of the material for such an analysis is now available through the ABS household expenditure survey and the subsequent fiscal incidence study, and in Jordan's Common Treasury, but we are still far from having a good picture of the comparison.  One complexity which needs to be taken into account is the existence of assets, for many elderly will be asset-rich even if income-poor, while young families may be asset-poor even if income-rich.  It is also important to remember that any prudent person has a longer time-span in which to prepare for the costs of old age than he or she has to prepare for the costs of the child-rearing phase.  These questions will be reconsidered in Chapter Eleven;  they are mentioned here only as a suggestion for future research by others qualified to undertake it

---

ENDNOTES

1.  Common Treasury, 1, 1.

2.  Common Treasury, 23.

3.  Ibid. 10.

4.  Ibid. 12.

5.  Cass, Income Support for Families with Children, 24.

6.  Common Treasury, 2, 231.

7.  Ibid. 180f.

8.  Common Treasury, 2, 179.

9.  Ibid. 1, 75.

10.  Ibid. 2, 164.

11.  Ibid. 226.

12.  Nurick, Wealth and Power, 28f.

13.  Common Treasury, 2, 175;  Jordan's exact ratio is 158:207.

14.  Ibid. 171f;  the exact ratio is 192:244.

15.  Ibid. 1, 70.

16.  192:322;  calculated from the population proportions at ibid. 170f.

17.  Ibid. 179.

18.  Beggs & Chapman, Forgone Earnings from Child-Rearing in Australia.

19.  Espenshade, "The Price of Children and Socioeconomic Theories of Fertility".  See also his Investing in Children:  new estimates of parental expenditure.

20.  Whiteford, Issues in Assistance for Families -- Horizontal and Vertical Equity Considerations, 31.

21.  "Government Welfare Outlays:  Who Benefits?  Who Pays?".  Remarkably this evidence, though it is crucial to family assistance questions, and though it has been available since 1987, has nowhere been taken into account or even mentioned by the Social Security Review.

22.  Common Treasury, 2, 189.

23.  EPAC Council Paper 35, 86.

24.  EPAC Council Paper 27, 38 & Chart 11.

25.  Oxley, The Structure of General Family Provision in Australia and Overseas, Table 1 (derived from OECD, The 1982 Tax/Benefit Position of a Typical Worker in OECD Member Countries.

26.  Oxley, op. cit., Table 4.

27.  The family assistance trend poses an interesting explanatory problem.  Goodin & LeGrand ("Creeping Universalism in the Australian Welfare State") provide four good reasons to expect universalism to flourish:  "boundary problems" in distinguishing poor from non-poor;  bureaucratic empire-building;  behavioural responses;  and political pressure.  Why have these not operated to expand universal assistance to families with dependent children?  Why, indeed, has family assistance steadily dwindled in this period?

28.  Edgar, "Volunteerism and the Changing Pattern of Women's Lives", 32.

29.  ABS Cat. 2506.0, 1989, 27-30.  These figures are a little out of date, and may need to be revised upwards by a few per cent.  See also the discussion of work-force participation in Chapter Eight below.

30.  Common Treasury, 1, 9,

31.  Australian Institute of Family Studies Newsletter 18 (1987), 23.

32.  Harding, Assistance for Families with Children and the Social Security Review;  Whiteford, Issues in Assistance for Families.

33.  EPAC Council Paper 29, 34;  Jordan, Common Treasury, 2, 190.

34.  Australian Institute of Family Studies Newsletter 18 (1987), 23.

35.  One -- who had spent ten years in the paid workforce -- is reported as saying "I have never experienced such emotional and physical exhaustion, or such little recognition for what I do on a daily basis":  Victorian Women's Consultative Council.  Women in the Home, 25.

36.  See Dixon, "How Mr Hawke has Cut Welfare Dependency", Table 2.  In 1988 there were 475,000 unemployment beneficiaries, 183,000 sole parent beneficiaries, 143,000 "widow" pensioners, as well as 388,000 invalid pensioners and 75,000 sickness beneficiaries (who -- presumably -- are unable to work).  Altogether, with their dependants, they would make up a population about the size of Melbourne.

37.  Berger, The War Over the Family, 172.

38.  See ABS Cat. 4119.0, 50-52.

39.  The War Over the Family, 175.

40.  Paul Amato's Children in Australian Families concluded that "a low level of [family] income is bound up with an unsupportive and fragmented family life".  Among young boys low family income "was associated with low support from mothers, low support from fathers, and low family cohesion".  These negative family features were strongest when income was very low.  Unfortunately, however, these correlations do not indicate in which direction the causation (if any) is running.  It may be that "an unsupportive and fragmented family life" is a cause of low income, not an effect of it.  Amato remarks that "Although governments can do little in the short-term to change parents' educational attainment, they can initiate policies to distribute more income towards families in poverty" (219).  The case for doing this would be very much stronger if we had a causal story which suggests that raising family incomes will improve family relationships which might in turn improve children's educational and occupational achievements.

41.  The War Over the Family, 182.

41.  In 1984, Chapman and Gruen (Unemployment, the Background to the Problem) estimated the total economic cost of unemployment at $4 billion.  Since then the unemployment rate has fallen but, with inflation, the cost is not likely to be less than $3 billion.