NATSEM

5426545689785426384512356458954526385745263685478954231 6478954265456897854263845123564589545263857452636854789 4231564789542654568978542638451235645895452638574526368 Financial 4789542315647895426545689785426384512356458954526385745 6368547895423156478954265456897854263845123564589545263 Disadvantage 5745263685478954231564789542654568978542638451235645895 5263857452636854789542315647895426545689785426384512356 5895452638574526368547895423156478954265456897854263845 in Australia 2356458954526385745263685478954231564789542654568978542 NATSEM 3845123564589545263857452636854789542315647895426545689 8542638451235645895452638574526368547895423156478954265 1990 to 2000 5689785426384512356458954526385745263685478954231564789 4265456897854263845123564589545263857452636854789542315 4789542654568978542638451235645895452638574526368547895 2315647895426545689785426384512356458954526385745263685 7895423156478954265456897854263845123564589545263857452 3685478954231564789542654568978542638451235645895452638 The persistence 7452636854789542315647895426545689785426384512356458954 of poverty in a 2638574526368547895423156478954265456897854263845123564 8954526385745263685478954231564789542654568978542638451 decade of growth 3564589545263857452636854789542315647895426545689785426 8451235645895452638574526368547895423156478954265456897 5426384512356458954526385745263685478954231564789542654 6897854263845123564589545263857452636854789542315647895 2654568978542638451235645895452638574526368547895423156 7895426545689785426384512356458954526385745263685478954 3156478954265456897854263845123564589545263857452636854 8954231564789542654568978542638451235645895452638574526 6854789542315647895426545689785426384512356458954526385 4526368547895423156478954265456897854263845123564589545 6385745263685478954231564789542654568978542638451235645 9545263857452636854789542315647895426545689785426384512 5645895452638574526368547895423156478954265456897854263 4512356458954526385745263685478954231564789542654568978 Ann Harding Rachel Lloyd everyone s family Harry Greenwell

The Smith Family s vision and mission statement In line with our vision for "a more caring and cohesive Australian community" The Smith Family researches different forms of disadvantage to propose preventive responses to them, and to promote social change through research, advocacy and innovation. Through advocacy we work for the development of social policy that benefits the entire community. Our mission is Together with caring Australians, to unlock opportunities for disadvantaged families to participate more fully in society. Author s note Ann Harding is Professor of Applied Economics and Social Policy and inaugural Director of NATSEM at the University of Canberra. Rachel Lloyd is a Senior Research Fellow and Harry Greenwell a Research Officer at NATSEM. General caveat NATSEM research findings are generally based on estimated characteristics of the population. Such estimates are usually derived from the application of microsimulation modelling techniques to microdata based on sample surveys. These estimates may be different from the actual characteristics of the population because of sampling and nonsampling errors in the microdata and because of the assumptions underlying the modelling techniques. The microdata do not contain any information that enables identification of the individuals or families to which they refer. ISBN 1 876833 02 5 The Smith Family, 2001 16 Larkin Street PO Box 10500 Camperdown NSW 1450 Telephone: (02) 9550 4422 Facsimile: (02) 9550 4235 www.smithfamily.com.au

Contents Author s note.............................................................. ii General caveat............................................................. ii Executive summary........................................................ vii PART A 1 Introduction......................................................... 2 2 Who is in poverty today?............................................... 4 3 Trends by family type................................................. 7 4 Main source of family income.......................................... 10 5 Labour force and educational characteristics............................... 15 6 Gender and age..................................................... 15 6.1 Gender......................................................... 15 6.2 Age........................................................... 17 7 Housing tenure and State of residence.................................... 18 7.1 Housing tenure................................................... 18 7.2 State........................................................... 20 8 How deep is poverty in Australia?....................................... 21 9 Conclusions........................................................ 23 PART B 10 Data and Methodology................................................ 26 10.1 An introduction to poverty measurement.............................. 26 10.2 Data issues..................................................... 31 PART C 11 Alternative poverty lines.............................................. 34 References......................................................... 39 iii

Ann Harding Inaugural Director of the National Centre for Social and Economic Modelling (NATSEM) and Professor of Applied Economics and Social Policy at the University of Canberra NATSEM is proud to once again be associated with The Smith Family's report on financial disadvantage and poverty in Australia. This is the second in a series of annual reports that aim to provide authoritative, up-to-date and freely available sources of information about poverty risk and depth in Australia. The National Centre for Social and Economic Modelling (NATSEM) was established in January 1993 at the University of Canberra. NATSEM has an international reputation as a centre of excellence for analysing microdata and constructing microsimulation models. It is a key contributor to social and economic policy debate and analysis in Australia, providing decision makers with the best available quantitative information to support their decisions. This series of reports complements NATSEM's ongoing research into income distribution and poverty in our society. We would like to thank Richard Percival, Gillian Beer and Simon Lambert for their assistance and Jan Gatenby and Keng Tan from the Australian Bureau of Statistics for their assistance in producing the 2000 results. Ann Harding iv

The Smith Family Elaine Henry Chief Executive Officer The Smith Family The Smith Family welcomes the release of the second annual Smith Family/NATSEM Report on Financial Disadvantage. We consider our continuing research partnership with NATSEM an important way to develop capacity to respond more effectively to the problem of financial disadvantage gaining evidence that can support directions for program and policy development to benefit all Australians. Strong economic growth over the past decade has not produced comparably strong social outcomes for many different groups within Australia. Indeed, the social impact has continued to raise concerns about the likelihood of achieving and maintaining adequate social and economic participation for significant portions of our population. This is a concern for those who understand the importance of bridging interests and opportunities among diverse groups to arrive at broad based agreement on values and activities central to the public wellbeing. While focusing on education as a key strategy in unlocking opportunities to stop people from falling into a cycle of poverty and in working for societal change, The Smith Family is also convinced of the importance of bringing about greater integration between economic and social policies for sustainable alternatives to persistent and recurring problems. The identification of trends over the past ten years in this year's report provides reliable indicators of how we can progress our work with others for bringing about such change. Elaine Henry v

Preface Dr Rob Simons National Manager, Research and Social Policy The Smith Family The Smith Family's Research and Social Policy Team has primary responsibility for the evaluation of our programs and social policy research. However, we also recognise the importance of collaboration in research and in providing evidence for the types of public policy that will enhance both social and economic outcomes. Hence, in some instances we commission research and enter into partnership with other researchers. The Smith Family's partnership with NATSEM has grown and been strengthened through our continuing collaboration in the production of this year's report Financial Disadvantage in Australia 1990 to 2000: The persistence of poverty in a decade of growth. It provides broad access to current data on financial disadvantage and best practice in its measurement. The subtitle of the Report, 'the persistence of poverty in a decade of growth', raises a number of questions, "Why did Australia not manage to make progress in the fight against poverty over the decade, despite falls in unemployment and steady economic growth during the period?"; "Why has the severity of poverty increased during the second half of the 1990's?"; and "Why do the majority of poverty measures suggest that Australia has not made any progress in the fight against poverty?" Such questions challenge us to re-examine our present policies and to develop broad cross-sectoral research collaborations to underpin more effective interventions and public policy. The Report reaffirms the relationship between education and financial disadvantage. Poverty rates among those aged 15 years and over decline sharply as educational qualifications increase. The poverty risk among those with university education is less than one half for those with no post secondary qualifications. While graduates have remained relatively immune from rising poverty in the 1990s, those with no secondary school qualifications have seen their poverty risk increase from 12.1 to 14.7%. These data support The Smith Family's focus on education for capacity building, strongly re-affirmed in a meeting of our Council of Governing Members, Board of Directors, and Senior Management earlier this year. The Smith Family is convinced that one of the key ways we can tackle some of the difficult questions raised by this year's Report, is to continue conducting research that provides further data on the relationship between education and financial disadvantage. In addition to research on factors that influence the educational performance of students from disadvantaged backgrounds, we are also continuing to look at the role of education in bridging the gap between the 'haves and the have nots'. An examination of the school to adult life transition strategies of Learning for Life students, and ways to improve the retention rates and academic outcomes of indigenous students are further projects through which we are collaborating with other researchers to find ways to move beyond some of the persistent problems that accompany financial disadvantage. Dr Rob Simons vi

Executive Summary This report provides estimates of trends in poverty in Australia over the past decade. There is no consensus among researchers as to the best measure of poverty. In this report we have used a common measure of relative poverty, setting the poverty line at half the average family income of all Australians. In 2000, this poverty line for a single income couple with two children was $416 a week. This is after the payment of income tax and before the payment of housing costs. In 2000, about one in every eight Australians lived in income poverty 13 per cent of all Australians. The risk of being in poverty was higher for children than adults, with 14.9 per cent of children and 12.3 per cent of adults being in poverty. While progress was made in the first half of the 1990s in the fight against child poverty, these gains were fully eroded during the second half of the 1990s. Poverty among adults increased steadily during the decade. These changes led to a gradual increase in poverty among all Australians during the 1990s. Sole parents remain the group at most risk of being in poverty, although there has been a pronounced fall in their risk of being in poverty since the beginning of the decade. Single people, particularly young singles, have fared worst in the past decade with almost one in every five single people now being in poverty. There has been a slight increase in the poverty rate among couples with children, and it is this increase that underlies the minor rise in child poverty during the 1990s. Government cash benefits are now the main family income source for almost six out of ten people in poverty. The poverty rate among those relying on government benefits has increased sharply, from 24 per cent at the start of the decade to 31 per cent by the end of the decade. More than half of all Australians who are unemployed live in a family that is poor. While the growth of part-time, low paid and casual work has prompted concerns about the possible rise of working poverty, there has been little change in the poverty risk faced by families whose main income source is wages and salaries during the past decade. About 15 per cent of all Australians in poverty in 2000 lived in families whose main income source is wages and salaries, much the same as in 1990. While full-time employment generally protects a family from being in poverty, part-time employment is less likely to do so. With the growth of part-time employment, part-time workers make up a growing share of the poor. Perhaps defying conventional wisdom, the poverty rate among women is now lower than for men. While the poverty risk for both men and women aged 15 years and over was exactly the same at the beginning of the 1990s, by 2000 men faced a slightly higher risk of being in poverty. This appears to be due to the sharp falls in poverty rates among female sole parents and the rising poverty rates among single males. Single males have fared much worse than single females as changes to industry structure and labour market conditions have impacted more harshly upon them during the past decade. Poverty rates generally decrease with age, although for a large part of the decade poverty among those in the 50 to 64 year age bracket was higher than for people aged 25 to 49 or over 65. The key change during the past decade has been the steady increase in poverty rates among the aged. While the risk of being in poverty used to be much lower for the aged than for those of workforce age, now the risk is relatively similar. A rather different picture is gained, however, if poverty is assessed after housing costs have been met. Because owner-occupiers have relatively low housing costs, their poverty rates fall as the focus shifts from a before to an after-housing picture. Conversely, the poverty rates of home purchasers and renters increase. The aged have high rates of outright home ownership and their risk of being in poverty has not changed during the 1990s if poverty is assessed after housing costs have been paid. On the other hand, poverty among adults aged 25 to 44 years has become more serious, with almost one-fifth of adults aged 25 to 44 years being in poverty on an after-housing basis. vii

The results by State fluctuate, and it is not clear whether the year-to-year changes reflect real variation or sampling error for the smaller states. NSW had the highest poverty rate in 2000, with 13.9 per cent of its residents being in poverty. Poverty has also increased in the other States, with only Queensland and Tasmania remaining relatively unchanged. The depth of poverty does not appear to have changed greatly during the 1990s, with the average gap between a poor family's income and the poverty line increasing from $112 in 1990 to $118 in 2000 (after taking out the impact of inflation). However, it appears that the severity of poverty increased in the second half of the 1990s. There is no agreement about where to set the poverty line and what equivalence scale to use. In Part C we present results for 12 possible poverty lines. In 11 of these 12 cases poverty rates increased during the 1990s, while in the remaining case poverty was stable. It thus appears that Australia did not manage to make any progress in the fight against poverty, despite falls in unemployment during the period. viii

Part A 1 Introduction......................................................... 2 2 Who is in poverty today?............................................... 4 3 Trends by family type................................................. 7 4 Main source of family income.......................................... 10 5 Labour force and educational characteristics............................... 15 6 Gender and age..................................................... 15 6.1 Gender......................................................... 15 6.2 Age........................................................... 17 7 Housing tenure and State of residence.................................... 18 7.1 Housing tenure................................................... 18 7.2 State........................................................... 20 8 How deep is poverty in Australia?....................................... 21 9 Conclusions........................................................ 23

Part A 1. Introduction This is the second edition of an annual comprehensive report on financial disadvantage among Australians. The key goal of this annual report is to provide authoritative and up-to-date estimates of financial disadvantage (or 'poverty') in Australia. While last year's report (Harding and Szukalska 2000a) primarily focussed on poverty during a single year, this year's report provides estimates of trends in poverty over the past decade. The concept of 'poverty' is vague and consequently the measurement of poverty is far from straightforward. There has been much discussion, over many years, about how best to define or characterise poverty (see Saunders 1994, pp. 221-30 and Osberg 2000, pp. 3-7 for recent reviews of the debate). Typically, poverty is regarded as a state of deprivation, a situation where one's standard of living has fallen below some acceptable minimum level. Absolute poverty occurs when families do not have sufficient income to pay for such basic necessities as food and housing. However, most poverty studies in developed countries have focussed on relative poverty, where a family's income is low relative to that of other families (for Australian examples, see the Commission of Inquiry into Poverty 1975; Johnson 1987; King 1997; ABS 1998). Income poverty lines are levels of income. If a family's income (adjusted to take account of the family's size and structure) falls below the poverty line then the family is considered to be in poverty. In the main part of this study, we have used a common measure of relative poverty, setting the poverty line at half the average family income of all people in Australia. (This measure does not adjust income to take account of housing costs.) As a measure of relative poverty, the half average poverty line has the advantage of being relatively transparent, as it is explicitly linked to the income distribution (Osberg 2000, p. 4). It can also be updated in a natural and consistent manner by calculating the average for each point in time. Some researchers prefer to set the poverty line at half the median (or middle) income of all people in Australia (ABS 1998). This implies a judgement that the living standards of the poor should be compared to those of people in the middle of the income distribution. The median also has the advantage of being less affected by extreme values than the average. For example, large increases in the highest incomes will cause the average to increase but alone will not have an effect on the median. However, in this report we have used the 'half average income' poverty line, as it better captures relative deprivation in times of rising income inequality (see Harding and Greenwell 2001). As the top income earners become better off, then the poor are relatively worse off. That is, they are poorer in relative terms. Using the 'half average income' line thus denotes acceptance of the proposition that the living standards of the poor should be measured against the living standards of all of the population, including the very affluent. The issues surrounding the way in which poverty is measured are discussed further in Part B. Measurement of poverty is also dependent upon the quality of data used. In this case, our data sources are the Income Survey confidentialised unit record files, released by the Australian Bureau of Statistics. Results are derived from surveys conducted in 1990, 1994-95, 1995-96, 1997-98, and 1999-00 but throughout the report the years are referred to as 1990, 1995, 1996, 1998 and 2000 for simplicity. There are some concerns about the comparability of the surveys conducted after the mid 1990s with that conducted in 1990. Again, these issues are explored in more detail in Part B. Because there is no consensus about the most appropriate poverty line to use, Part C provides estimates of poverty trends using a range of poverty lines, both before and after taking account of housing costs. In Part A, Section 2 examines trends in aggregate poverty between 1990 and 2000, while Section 3 undertakes further analysis by family type. Section 4 looks at whether there have been changes in the major income sources of families in poverty, while Section 5 explores the labour force and educational characteristics of families in poverty. Section 6 disaggregates the poverty picture by age 2

and gender, while Section 7 explores the impact of housing tenure and state of residence. Section 8 considers whether the depth of poverty in Australia has changed how far below the poverty line those families in poverty actually are. Finally, Section 9 summarises the findings of this report. In this report, we define dependent children to mean all children aged less than 15 and all 15 to 24 year olds engaged in full-time study and still living at home with their parents, in accordance with the ABS definition (ABS 1999, p. 68). Adults is defined as the residual. That is, the number of adults equals the total number of people in the Australian population minus the number of dependent children. The head of the family is the ABS 'reference person'. For readability we use the term family to mean the ABS income unit. What is Poverty? Part A It is hard to gain agreement about how poverty should be defined and measured in our society. For some people poverty means that a family cannot afford to buy food or adequate shelter. But in industrial economies like Australia a relative poverty definition is more commonly used. In this study we have set the main poverty line at half the average family income of all Australians. Family income has been adjusted to take different family size and composition into account, using the Henderson equivalence scales. Using this method, we estimate that in 2000 the poverty line for a single income couple with two children was $416 a week. This means that the 'cash in the hand' of such a family has to be more than $416 a week for this family not to be in poverty. Cash income is measured as all income received from such sources as wages and investments, plus cash benefits from government such as the age pension, minus any income tax paid. This is called disposable income. The poverty line varies for different types of families, because larger families need more income to survive than smaller families. For example, for a single employed person the poverty line is $225 a week. 3

Part A 2. Who is in poverty today? The latest ABS data suggest that almost one in every eight Australians live in income poverty today, when the poverty line is set at half the average family income of all Australians (Figure 1). This represents 2.4 million people, or 13 per cent of the population. Last year's report showed that the risk of being in poverty is higher for children than adults, and the same result was evident in 2000 with 14.9 per cent of all dependent children in poverty compared with 12.3 per cent of adults. Has this picture changed over the past decade? Comparison with 1990 suggests that poverty has increased slightly over the 1990s (Table 1). For children, there is a decline in poverty rates between the beginning and the middle of the decade and then an apparent increase in the second half of the decade. For adults, the picture is quite different and appears to be one of consistent but slow increase in the risk of poverty. Because adults so greatly outnumber children, the pattern for adults prevails and results in the slow but steady increase in poverty rates among all Australians over the 1990s (Figure 2). Figure 1 Estimated number of Australians in poverty in 2000 a 2,432,000 or 13 per cent Adults 1,688,000 (poverty rate 12.3%) Dependent Children b 743,000 (poverty rate 14.9%) Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). See p.29 for further explanation. b) Dependent children means all children aged less than 15 and all 15 to 24 year olds engaged in full-time study and still living at home. Figure 2 Estimated poverty rates, 1990 to 2000 a poverty rate % Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). 4

Table 1 Estimated poverty rates a and other characteristics of Australians, 1990 to 2000 b 1990 1995 1996 1998 2000 Half average income poverty line Poverty rates Children 14.3 13.3 13.1 13.8 14.9 Adults 10.4 11.0 11.5 12.2 12.3 All Australians 11.3 11.7 12.0 12.6 13.0 Number in poverty c '000 Children 511 633 627 684 743 Adults 1261 1427 1510 1633 1688 All Australians 1772 2060 2137 2317 2432 Real (inflation adjusted) equivalent disposable incomes of the: d Poor $224 $229 $233 $250 $262 Middle (median) $686 $669 $650 $693 $723 Average (mean) $766 $759 $747 $798 $832 95th percentile $1482 $1514 $1490 $1611 $1654 Part A Ratio between incomes of the: Middle/poor 3.06 2.92 2.79 2.77 2.76 Average/poor 3.42 3.31 3.20 3.20 3.18 Half median income poverty line Poverty rates Children 9.0 8.4 8.7 8.5 9.6 Adults 8.0 8.2 8.5 8.4 8.4 All Australians 8.2 8.2 8.6 8.4 8.7 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). b) Results are derived from surveys conducted in 1990, 1994-95, 1995-96, 1997-98, 1999-00 but throughout the report the years are referred to as 1990, 1995, 1996, 1998 and 2000 for simplicity. c) The increasing number of people in poverty is due to two factors the increasing proportion of people in poverty and the increasing population. d) These are 'equivalent' incomes for all families (weighted by the number of people). 'Equivalent' incomes are adjusted for the needs of the family, and vary with the number and labour force status of the adults and the number and age of children. All dollars in this report are 2000 dollars, with values for earlier years inflated by the Consumer Price Index. The 'poor' here means those defined as poor using the half average income poverty line, and the figures given are the average incomes for all poor people. The 'middle' means the median family income and the 'average' indicates the mean family income. These results suggest a discernable increase in poverty over the 1990s, when using the 'half average income' poverty line. But has this been due to a decline in the incomes of the poor? Or to strong economic growth lifting average incomes for those in work and thus the poverty line? Or is movement at the top of the income distribution affecting the average income and thus the poverty line? Figure 3 indicates that the average incomes of poor families, after taking out the impact of inflation, have increased by $38 a week during the 1990s. How does this growth compare with the incomes of non-poor families? Figure 3 also shows the change in median (middle) family incomes over this period, and suggests that there has also been a slight increase in median family incomes, of $37 a week between 1990 and 2000. Finally, Figure 3 traces the change in average Australian incomes over the past two decades. It shows that average 5

Part A incomes have increased by $66 a week almost twice as much as for middle and poor families. As a result of these differential growth rates, the gap between middle and average incomes has been growing, up from $80 in 1990 to $109 in 2000. While average incomes have increased strongly, this growth appears to have been primarily driven by increases at the top end of the income distribution, as shown by the rising average real incomes of those at the 95th percentile of the income distribution indicated in Table 1. The incomes of those at the 95th percentile have increased by $172 a week during the 1990s. The 1990s have thus seen a widening gap between the incomes of the poor and the incomes of the very affluent, rather than between the incomes of the poor and the incomes of middle Australian families. It is the rising incomes of the affluent that are generating much of the growth in the half average income poverty line, and thus the increase in poverty rates apparent in the 1990s. This divergence between the increases in middle family incomes and incomes at the top end means that a somewhat different impression of poverty trends is gained if the poverty line is set at half of middle family income rather than half of average family income. If the poverty line is set at half median income rather than half average income, poverty has remained relatively stable over the 1990s. As the lower shaded rows in Table 1 indicate, using a half median income poverty line results in poverty remaining at 8.2 per cent in 1990 and 1995, and then rising to 8.7 per cent in 2000. Irrespective of the poverty line used, both sets of results indicate that Australia has not been successful in reducing poverty in the 1990s, despite falling unemployment. Figure 3 Trends in poor, middle, average and top incomes, after adjustment for inflation, 1990 to 2000 a dollars per week Notes: a) All figures are for equivalent disposable family income, adjusted by the CPI to 2000 dollars. 'Poor' families are those in poverty using the Henderson half average income poverty line. 6

3. Trends by family type The type of family that individuals live in has a major impact upon their likelihood of being in poverty. Being part of a couple offers some protection against poverty, as there is another potential income earner in the family and costs such as housing are shared. Those living in sole parent families continue to face the highest risk of poverty out of the four broad family groupings considered in Table 2, with more than one in every five Australians living in sole parent families being in poverty. Single people are the group facing the second highest risk of poverty. An estimated 18.3 per cent of all single people are in poverty in Australia in 2000, with the risk varying markedly with age. Young singles aged 15 to 24 years experience higher poverty rates than singles generally (20.8 per cent versus 18.3 per cent). This is partly because young people generally have lower incomes as they are at the start of their careers or fill low paid jobs while they are studying. The young also face high unemployment rates and, as detailed later in the report, poverty among the unemployed is very high. However, this apparently high youth poverty rate has to be treated with some caution, as about half of all single 15-24 year olds in poverty are non-dependent children still living at home with their parents. While it cannot be assumed that their parents are always willing and able to offer financial support, in many cases such support probably extends to free or subsidised housing, meals and utility expenses. Those living in families consisting of a couple with children have about a one in eight chance of being in poverty (Table 2), with the risk increasing steadily as the number of children in the family increases. For example, those with three or more dependent children families are about twice as likely to be in poverty as those with only one child 19 per cent versus 8.6 per cent (Figure 4). Finally, about one in every 16 people living in 'couple without children' families are living below the poverty line. Part A Table 2 Estimates of poverty rates of individuals, by family type, 1990 to 2000 a Family Type 1990 1995 1996 1998 2000 Sole parent 28.0 21.3 20.8 21.4 21.8 Single person 13.7 15.0 15.4 16.5 18.3 Couples with children 10.4 11.3 11.1 11.1 12.2 Couples without children 6.6 6.7 7.9 9.0 6.4 All family types 11.3 11.7 12.0 12.6 13.0 a) Using the before-housing half average income poverty line (Henderson equivalence scale). Has there been a change in the poverty risks faced by different family types over the last decade? Figure 5 shows that the most striking change has been the dramatic fall in poverty among sole parent families, with other research suggesting that this has been due both to improvements in government assistance and increases in child support resulting from the introduction of the Child Support Scheme in 1988 (Harding and Szukalska 2000b). The poverty risk among couples with children has risen by about one-fifth during the 1990s. Single people have fared the worst in the past decade. The poverty rate for single people has moved up from 13.7 per cent in 1990 to 18.3 per cent in 2000. The picture for couples without children is less clear, with a strong increase in the poverty rate from 1990 to 1998 but a lower rate in 2000. Those living in sole parent families face a very high risk of being in poverty. But because relatively few Australians live in this type of family, they make up only a small proportion of all Australians in 7

8 Part Part A poverty. On the other hand, couples with children face the same risk of being in poverty as Australians generally, but so many Australians live in this type of family that they make up the single largest family grouping among those in poverty. About 42 per cent of all Australians in poverty live in families which consist of a couple with children, while another 13 per cent live in sole parent families (Figure 6). Just over one-third of all those in poverty are single people, while another 12 per cent are couples without children. Figure 4 Estimated poverty rates for individuals living in sole parent and couple with children families, by number of children a Sole parents Couples 25.9 poverty rate % 15.4 19.0 8.6 9.4 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). Figure 5 Estimated trends in poverty rates of individuals, classified by family type, 1990 to 2000 a poverty rate % 1990 1995 1996 1998 2000 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale).

Figure 6 Estimates of Australians in poverty, by family type, 2000 a Couples with children 42% Single person 33% Part A Couples without children 12% Sole parent 13% Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). 9

Part A 4. Main source of family income For almost six in ten people in poverty, government cash benefits such as the age pension or unemployment allowances are their family's main source of income (second panel in Table 3). Previous research has shown that the main types of poor families that receive government cash payments are the unemployed (Harding and Szukalska 2000a, p. 5). Generally, pensioners who are not in the labour force receive cash benefits from government that lift them above the poverty line. Among those in poverty there is also a small group whose main income source is family allowances, with the value of these payments exceeding their very low incomes from other sources such as selfemployment or casual work. Table 3 Estimates of poverty among individuals, by main income source of family, 1990 to 2000 a 1990 1995 1996 1998 2000 Poverty rates Zero income 100 100 100 100 100 Government cash benefits 23.9 24.9 24.8 25.0 31.1 Other income sources 25.3 22.7 26.8 23.8 20.3 Own business 14.7 12.9 14.6 17.4 14.3 Wage and salary 2.9 3.4 3.2 3.5 3.2 Proportion of those in poverty Government cash benefits 46 52 51 50 58 Wage and salary 16 18 16 17 15 Other income sources 17 11 13 12 9 Zero income 11 11 9 11 9 Own business 10 8 11 10 8 Total 100 100 100 100 100 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). Figure 7 Australians in poverty by main source of family income, 1990 to 2000 a proportion of those in poverty Other income sources (such as investment and superannuation income) Own business Zero income Wage and salary Government cash benefits Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). 10

The second most numerous group is the 'working poor' families whose main income source is wages and salaries but whose incomes are below the poverty line. 1 The risk of being in poverty among wage and salary earners is fairly low (only 3.2 per cent in 2000) but, because wage and salary earners comprise such a large share of Australian families, this small rate still represents a significant number of people. In 2000 an estimated 15 per cent of all those below the poverty line had wages and salaries as their main family income source. The remaining 26 per cent are split relatively evenly between families that have zero incomes 2, business income or other income sources (such as superannuation, dividends or rent from an investment property) as their primary income source. Have there been major changes in the primary income sources of poor families over the past decade? The most evident trend is the increasing poverty rate for people in families with government cash benefits as their main income source, which rises from 24 per cent in 1990 to 31 per cent in 2000. This group also made up a growing share of the poor 58 per cent in 2000, up from 46 per cent in 1990 (Figure 7). The other key trend is that poverty among people in families with 'other income' as their main income source has decreased. However, this is a fairly small group, so sample size and comparability of the surveys may have influenced this result. Part A 1. There is no accepted definition of the 'working poor'. For an alternative approach to the study of working poverty, see Eardley (1998). 2. For some years the income survey data have recorded negative business and investment incomes (for example, this might be the result of negative gearing on rental properties). However, to be comparable with 1990 data these have been reset to zero. The group with zero incomes might also include those with no earned income and assets that exceed the asset test limits for income support payments, such as the aged pension. There is also a group of people who report that they have no income on the survey, despite the fact that their other characteristics such as employment type, labour force status and age would suggest that they might be eligible for income support payments from the government. 11

Part A 5. Labour force and educational characteristics Having a job has long been regarded as one of the key ways to protect a family from being in poverty. This premise is confirmed in Table 4, which looks at the poverty rates of all people aged 15 years and over by their labour force status. It suggests that only 4.6 per cent of Australians who hold a full-time job live in families that are in poverty, with the poverty risk increasing to 11.7 per cent among Australians aged 15 and over who are working part-time. More than half of all Australians who are unemployed live in a family that is poor, underlining the major relationship between unemployment and poverty. Finally, about 17 per cent of Australians not in the labour force live in poor families. Table 4 indicates that poverty rates for the unemployed have increased sharply during the second half of the 1990s. This may be a result of government policy change initiated in early 1998. Prior to this, the rate of allowances (such as the unemployment allowance) for couples was the same as the rate of pension for couples. However, since 1998, unemployment allowances have been indexed to the Consumer Price Index, while pensions have been indexed to movements in average earnings. As a result, while the rate of payment received by an allowee couple was the same as that received by a pensioner couple in 1997, now the allowee couple receive $27.40 per fortnight less than the pensioner couple. Those receiving unemployment allowances have thus fallen further behind average incomes within the community. Table 4 Estimated poverty rates by labour force status for persons aged 15 and over a 1990 1995 1996 1998 2000 Unemployed 52.6 49.2 47.6 54.5 57.5 Not in the labour force 13.4 12.4 13.4 14.2 16.7 Working part-time 10.7 10.7 12.7 13.0 11.7 Working full-time 4.2 5.2 5.5 5.3 4.6 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). Table 5 shows that the risk of being in poverty declines sharply as the number of income earners in the family increases. These results differ from those in Table 4, as they also include those aged less than 15 years. For example, in 2000, 28 per cent of people in families with no income earner and 24 per cent of people with only one part-time earner in the family were in poverty. In contrast, just under seven per cent of people in families with one full-time earner are in poverty. Again echoing the findings in Table 4, full-time work is thus more likely to pull a family out of poverty than part-time work. Despite the low poverty risk among people living in families with a full-time earner, the second panel of Table 5 shows that of people in poverty, 15 per cent live in a family with one full-time earner and 5 per cent live in a family with 2 full-time earners. 12

Table 5 Estimated poverty among individuals, by number of earners in their family, 1990 to 2000 a 1990 1995 1996 1998 2000 All families (including wage and salary earners and self-employed) Poverty rates No earners 25.5 24.4 23.7 25.5 28.2 One part-time earner 21.2 18.5 19.1 25.8 24.2 One full-time earner 5.3 6.9 5.9 5.4 6.5 Two earners (at least one part-time) 6.7 7.1 9.6 7.6 7.8 Two full-time earners 4.4 4.8 5.4 6.2 4.0 Proportion of those in poverty b No earners 57 55 51 54 55 One part-time earner 9 8 10 13 14 One full-time earner 16 19 15 13 15 Two earners (at least one part-time) 11 10 14 10 11 Two full-time earners 6 6 7 7 5 Part A Families with wage and salary earners only Poverty rates One part-time earner 19.4 16.2 13.8 22.3 21.9 One full-time earner 1.9 2.9 1.5 2.1 2.4 Two earners (at least one PT) 3.2 4.5 6.8 3.9 4.3 Two full-time or three earners 1.4 1.7 2.0 1.9 1.6 Proportion of those in poverty b One part-time earner 39 28 29 44 46 One full-time earner 28 33 15 21 21 Two earners (at least one PT) 23 26 41 23 23 Two full-time or three earners 10 13 15 13 10 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). The number of earners includes the number of dependent children earners as well as parental earners. b) The proportions do not sum to exactly 100 per cent as there are a small number of families with three income earners (i.e. both parents are working and so is a dependent full-time student). The top half of Table 5 includes the self-employed, whose incomes may not always accurately reflect their real financial position (Bradbury 1996). To deal with this, the bottom half of Table 5 looks just at families with some wage and salary income. The risk of being in poverty is dramatically lower among wage and salary earners working full-time than among the self-employed. While families with one part-time wage and salary earner face a 22 per cent risk of being in poverty, those with one fulltime wage and salary earner face only a two per cent risk and those with two full-time wage and salary earners face an even lower 1.6 per cent risk. It thus appears that for the overwhelming majority of Australians, a full-time wage and salary job is sufficient to pull a family out of poverty. There have been suggestions that having a job may no longer be as effective a guarantee against poverty as in the past, given the rise in part-time and casual work and changes in earnings patterns. Table 4 appears to offer some limited support to this proposition, with the poverty risk among those Australians aged 15 and over working either full-time or part-time being slightly higher at the end of the 1990s than at the beginning. That is, while 10.7 per cent of all Australians working part-time were in poverty in 1990, by 2000 this had increased to 11.7 per cent. Similarly, looking just at the wage and salary earning families in the bottom panel of Table 5, there is a marginal increase in the 13

Part A risk of being in poverty for all four of the earnings categories considered. Overall, however, the major increases in poverty risk appear to be among those who are unemployed or not in the labour force, rather than among those who are employed. While having a job provides significant protection against poverty, so too does higher education. As Table 6 indicates, poverty rates among those aged 15 and over decline sharply as educational qualifications increase, with the poverty risk among those with university education being less than half that for those with no post-secondary qualifications. While graduates have remained relatively immune from rising poverty in the 1990s, those with no post-secondary school qualifications have seen their poverty risk increase from 12.1 to 14.7 per cent. Table 6 Estimated poverty rates by highest educational qualification for persons aged 15 and over a 1990 1995 1996 1998 2000 No post secondary qualifications 12.1 12.2 13.0 13.6 14.7 Still at school 14.2 11.9 16.0 16.1 14.0 Diploma, certificate, trade qualifications 8.7 9.8 10.2 10.6 10.5 Bachelor degree or higher 6.0 7.4 6.8 8.1 6.0 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). 14

6. Gender and age 6.1 Gender One question of enduring interest is the extent to which poverty has a female face (Cass and O'Loughlin 1984). In last year's report we focussed on the gender of the head of the household, the results for which are dominated by men in couple families and women in sole parent families. However, in Table 7, for the first time, we have estimated the poverty rates for men and women as individuals. As the ABS does not record the gender of children in its national income surveys, estimates of poverty by gender for the whole population cannot be calculated. The following analysis is thus only for those aged 15 years and over. Because of women's greater likelihood of being sole parents (Lambert 1994) and their lower earnings when in the labour force (ABS 2000) it might be expected that more women would be in poverty than men. The results in Table 7 are thus a surprise, suggesting that the risk of being in poverty for women is slightly less than it is for men 12.0 per cent compared with 12.5 per cent. Part A Table 7 Estimates of Australians aged 15 years and over in poverty, by gender, 1990 to 2000 a 1990 1995 1996 1998 2000 Gender Poverty rates Males 10.6 11.0 11.8 12.3 12.5 Females 10.6 10.9 11.5 12.1 12.0 Proportion of those in poverty Males 49 50 50 50 50 Females 51 50 50 50 50 TOTAL 100.0 100.0 100.0 100.0 100.0 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). Has there been a change in the risk of poverty by gender in the last decade? In 1990, the poverty rates for men and women were the same and both experienced increases up to 2000. However, the poverty rate for men has increased faster than for women. Why do we have such a surprising result? The group known to have contributed to a high female poverty rate sole parents have experienced sharp falls in the risk of poverty and this has been one factor acting to reduce female poverty. In addition, the rising tide of poverty among single people younger than retirement age seems to be impacting more on men than women. Table 8 shows the number in poverty and the poverty rates of men and women by family type. There are considerably more poor women in sole parent families than there are men (104 000 compared with 34 000) and there are 106 000 poor single women over 65 as compared with 40 000 men in this group. However, the number of poor single men under 65 is 409 000, compared with just 259 000 poor single women. It is this gap that is giving rise to the higher male poverty rates by the end of the decade. 15

Part A Table 8 Estimates of number of Australians aged 15 years and over in poverty, by gender and family type, 1990 to 2000 a 1990 1995 1996 1998 2000 '000 '000 '000 '000 '000 Males Couples with dependents 220 243 261 261 278 Couples without dependents 138 147 172 198 148 Sole parents 29 27 34 26 34 Singles 15-24 151 153 168 183 164 Singles 25-64 125 148 163 190 245 Singles 65+ 15 33 15 18 40 Females Couples with dependents 217 253 253 251 283 Couples without dependents 138 147 172 198 148 Sole parents 93 81 81 112 104 Singles 15-24 121 132 141 140 102 Singles 25-64 86 109 117 115 157 Singles 65+ 44 39 49 69 106 Notes: a) Using the before-housing half average income poverty line (Henderson equivalence scale). An analysis of unemployment rates for men and women shows that, until the early 1990s, the unemployment rate for women was higher than for men but since that time the unemployment rate for women has been consistently lower (Figure 8). The difference between the unemployment rates for single men and women is even more pronounced. The decline in manufacturing industries (that have a high proportion of male workers) and the rise of service industries (that have more female workers), coupled with the rise of part-time and casual work (which tends to be dominated by women), are likely to be factors that have impacted on the new trends. Figure 8 Unemployment rates by gender, 1983 to 2001 Unemployment Rates % 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 Source: ABS AusStats Service, Labour Force (UR) Unemployment Rate, Cat. No. 6291.0.40.001 16