Chapter 2 - Defining and measuring poverty

Chapter 2 - Defining and measuring poverty

Poverty is essentially the lack of the means to live. At the heart of any consideration of poverty lies the issue of what is needed to live "a decent life" and, more fundamentally, what it is to be human.[1]

2.1        This chapter provides an overview of issues relating to poverty and the measurement of poverty. Specifically, it addresses issues related to the differing definitions of poverty, the ways in which poverty may be measured and various problems that arise in any attempt to measure poverty. These issues have been the subject of extensive debate and controversy over many decades. That this is the case is not surprising. As one submission noted 'in effect the measurement of poverty and inequality seeks to quantify a set of values, and value judgements, of a society'.[2]

What is poverty?

2.2        Poverty as a concept is difficult to define and during the Committee's inquiry there were differing views amongst participants as to what constitutes poverty or how best to measure it (a problem shared with academic researchers and the community generally). One study has argued that the only point of general agreement is that people who live in poverty must live in a state of deprivation, a condition in which their standard of living falls below some minimum acceptable standard.[3]

2.3        Submissions and other evidence to the inquiry provided a range of approaches to the conceptualisation and definition of poverty. Poverty can be broadly defined in absolute or relative terms. Absolute poverty refers to people who lack the most basic of life's requirements and is measured by estimating the numbers of individuals or families who cannot provide for the necessities of life such as housing, food or clothing. Commentators have generally argued that the meaning of poverty in a relatively advanced country like Australia is quite different from the absolute deprivation or subsistence which exists in many developing countries and therefore the concept of absolute poverty has little relevance to conditions in Australia.[4] However, the Aboriginal and Torres Strait Islander Commission (ATSIC) noted that some remote Indigenous communities are in absolute poverty in Australia, measured by poor infrastructure with associated diseases that are largely eradicated in other parts of Australia and with high child mortality rates.[5] Another submission also commented that instances of infant mortality and malnutrition amongst Indigenous Australians, for example, indicate the presence of absolute poverty in Australia.[6] Other examples of absolute poverty include the homeless or those at risk of homelessness and/or those with very few material possessions who rely on social security benefits and charities just to 'get by'.

2.4        In affluent Western countries like Australia poverty is usually conceived in relative rather than absolute terms. Relative poverty refers to individuals or families that have low incomes or other resources relative to other individuals or families. Relative poverty is defined not in terms of a lack of sufficient resources to meet basic needs, but rather as lacking the resources required to participate in the lifestyle and consumption patterns enjoyed by others in the society.[7]

2.5        Evidence to the Committee has generally defined poverty in relative terms as is illustrated in the definitions provided below.

Definitions of Poverty


Poverty is an enforced lack of socially perceived necessities. This definition has three core elements:

Submission 163, p.38 (ACOSS).

St Vincent de Paul National Council

Poverty has two elements:

While financial disadvantage is an important part of the problems for the poor and disadvantaged, equally important is the inequality of opportunity. It is this lack of opportunity which brings about poverty and/or prevents people from escaping the poverty cycle.

Submission 44, p.6 (St Vincent de Paul National Council).

Mission Australia

Poverty includes social, psychological and spiritual dimensions, in addition to financial hardship.

The inevitable consequence of adopting such an inclusive understanding of the nature of poverty is to appreciate the breadth of the impact which poverty has on a person's ability to function as a full member of society. It is not just a matter of being unable to buy things, but of being unable to participate – of being excluded and isolated from one's fellow human beings.

Submission 169, p.8 (Mission Australia).

UnitingCare Australia

Poverty is essentially the lack of the means to live.

The Christian tradition understands that people have, in addition to basic physical needs such as food, shelter and clothing, other basic needs, such as the need for education, the need to participate in society and contribute to the common good, the need for intellectual, cultural and creative activity, the need to participate in religious activity and community and the need for rest and recreation. Without these other basic needs being met, human beings may survive but do not flourish.

Submission 133, p.3 (UnitingCare Australia).

Brotherhood of St Laurence

Poverty is both:

"Adequacy" is defined in relation to community standards and may change over time, as do standards of living for the community as a whole.

Submission 98, p.1 (Brotherhood of St Laurence).

2.6        While definitions of poverty vary, they generally share common features. As noted above, ACOSS defines poverty as an enforced lack of socially perceived necessities, which has a number of core elements which it shares with other definitions of poverty.

2.7        Necessities are resources – goods and services – that people cannot reasonably live without. The adequacy of resources and whether or not they are necessary needs to be gauged according to people's needs. Poverty is usually defined with reference to a set of basic human needs – physical survival and comfort, and the need to participate in the economic and social life of the community. This suggests that poverty has two forms which ACOSS refers to as 'subsistence poverty' – an inability to meet basic physical needs, and 'participation or social poverty' – an inability to meet social needs. Examples of subsistence poverty include inadequate diet, sub-standard housing or denial of basic health services. Examples of participation poverty include an inability to meet regularly with family or friends, to travel to job interviews, to afford school books and excursions for children, or to complete an education. These forms of poverty are closely related but are distinct. People may have all the resources they need to subsist but lack the resources required for social participation. Participation poverty has much in common with the idea of social exclusion, a term that usually refers to the exclusion of people from mainstream social and economic life.[8]

2.8        Similar arguments were provided by Professor Peter Saunders of the Social Policy Research Centre who argued that despite the apparent diversity in the definitions of poverty most encompass a number of core features. These are that poverty is a situation in which resources are not adequate to meet basic needs and that any definitions usually embody community perceptions of poverty in some way. He argued that a definition of poverty as an enforced lack of socially perceived necessities captures the critical aspects of poverty succinctly. It also emphasises the fact that poverty is a situation which is forced onto people, not chosen by them.[9] This fundamental point was constantly raised in evidence to the inquiry.

2.9        Several submissions argued that the lived experience of the poor should also be taken into account in defining poverty. One submission noted that ideally submissions to a poverty inquiry such as this inquiry should not be written by professional social analysts, economists or even welfare practitioners but should be written by the poor themselves  – 'only those who are, or have been, genuinely poor can do justice to the experience of the suffering of those economically disadvantaged'.[10] Another submission referred to a number of social surveys of disadvantaged people which demonstrated their capacity to present a detailed and multifaceted understanding of what it means to be disadvantaged in Australian society.[11]

2.10    The Committee recognises these issues and was deeply moved by the personal descriptions from those people experiencing poverty and those working directly with the poor and disadvantaged 'at the coalface'. The Committee's many hearings conducted throughout urban and regional Australia provided an opportunity to hear directly of the experiences and difficulties of people living in poverty and those working closely with them, especially through welfare agencies. Their collective insights were particularly valuable and provided a useful reference point for framing many of the Committee's approaches to poverty alleviation. A number of these individual contributions are provided below.

What it Means to be Poor

A lot of people that have been here have spoken of how they came to be in poverty. I am speaking as someone who lives in poverty...It does not matter how a person comes to live in poverty, and it does not matter what it says in the dictionary as its definition; poverty is everything. You cannot afford basic needs. The first thing that you let go of is yourself...When you are sitting down trying to work out a budget on $342, straightaway you see $48 for groceries and $28 for toiletries and other necessities – that is $70 already. Then you add rent on to that. That is not having a car – which is a luxury, I believe – or a telephone, because it is not a necessity. The needs of life are not covered by what is put in. It does not matter whether you are on unemployment or a pension. You break it down: it is not enough for anybody to sustain a standard of living.

Committee Hansard 4.08.03, p.1198 (Miss Catherine Gammage, Brisbane).

I live in a moderately middle-class suburb...It is not totally affluent but I know some people around me in the cul-de-sac and up the street who cannot afford heating. They go to bed at dark. They cook their dinner early, they turn on their electric blanket and they take their dinner to bed with them so that they will not have to turn on the heating. They cannot afford wood. They cannot afford their oil heater and their oil heaters are sitting there unused and going rusty. That is not just one or two people; there are many. People say, "What a good idea! Right." So for quite a number of people that is the norm in what is not a poor suburb. These are people who have worked hard all their lives. They are not on pensions, they are not eligible for a pension, but they cannot afford heating.

Committee Hansard 2.05.03, p.222 (Tasmanian Poverty Coalition).

What I hear is absolute pain, the pain of poverty. Let us not walk away from that: it is a very painful experience. I am still haunted by the story of a person who did not appear at but whose story was told at a recent Just Jobs conference. This was a man who lived in the Huon Valley who had become totally isolated by his poverty, to the degree that he did not leave or very rarely left his home. He remained shut up in his house because of the shame, the feelings of rejection and the sense of isolation from his community. When we have people who are shut up in their homes because of the experience of poverty, people who are not interacting with others, then I think we have a serious problem on our hands.

Committee Hansard 2.05.03, p.212 (Tasmanian Catholic Justice & Peace Commission).

Like millions of other low-income Australians, I am one of the hidden poor, just keeping afloat. We are flat out treading water out here. We are making very little headway towards our aspirations, and we are one crisis or catastrophe away from the poor box. We are living on the edge.

We live in the shadows of the dismal statistics. We are not mad, bad, sad or totally dysfunctionally overwhelmed by our life circumstances. Many of us are highly skilled and well educated. We are all doing what we can to contribute to society with the resources we have. Our poverty is poverty of resources, services and is getting too hard to make ends meet, let alone work towards our dreams.

Committee Hansard 1.07.03, p.874 (Miss Margaret Clarke, Byron Bay).

I want to stress in relation to this delegation that we are not policy experts but we are experts on the lived experience – the lived experience of these people who have suffered the pain and heartache of poverty in the city of Sydney. It is clear from the stories we will hear today that poverty is on the increase. I have a clear example of this. I am a coordinator of our night patrol service. Here tonight in Sydney – at Martin Place, Kings Cross, Central Station and down the back of Kent Street ­ we work with homeless people. In 1998 we worked with 23,000 cases. In 2002 we worked with 43,000 cases. That is a 20,000 increase in the number of people we work with every year. I am not just talking about men with alcohol and drug addiction, I am talking about men, women and children. Only last Tuesday an unregistered Commodore car followed us around the back of Kings Cross. It was a family of three children and a single father, looking for something to eat and for some kind of support from us.

Committee Hansard 26.05.03, p.357 (St Vincent de Paul Society – Sydney Archdiocesan Council).

I move on to poverty. Personal family stress, constantly juggling finances or being in debt, no spare money to cover unexpected expenses, like the broken-down fridge or hot water service, even birthdays and celebrations, never being able to make lifestyle choices like going to the movies or going out for dinner. Because of those sorts of things people never develop a real sense of hope.

Committee Hansard 29.05.03, p.570 (Salvation Army, Newcastle).

My memory of one instance of real poverty, which I guess is why we are all here today, relates to a call for assistance we had from a client. We got a phone call from this fellow and I said, 'Before we come and see you, what's your problem?’ He said, 'I'm hungry' 'To save me the time, could you tell us what your basic requirements are?' This will stagger you. His answer was, 'Seven loaves of bread, some pasta, some vegemite, some cereal, three litres of milk, some meat sauce and a couple of incidentals.' We went and bought it. It cost $39.71.

So we went around to this chap's place in Pottsville, which is a pretty nice area to live, but it was a run-down place. It was I guess earmarked for redevelopment. We went inside with the client – and let us call him Trevor – and he was there with his two young boys. These kids were perhaps 10 and 12. We walked in with the bags of food and their eyes just lit up. We sat down at a battered camp table with four chairs – you know the fold-out gear – put the bags of food on the table and then started to talk. The kids straightaway got the bags of food, looked inside them and started to put the food away...So we were there and we were asking what else we could do for him. 'You've done absolutely more than enough,' he told us. 'I'll be right. There are more deserving people.' My companion on that day was the mayor of a well-known inland town in central New South Wales...he came away, like me, thinking that that was real poverty, just to see the way those kids operated. That guy had pride. He did not want any further help. He reckoned that was his help and that was going to get him over the line. They got all those loaves of bread, and you can imagine what they were going to do.

Committee Hansard 1.07.03, pp.851-52 (St Vincent de Paul Society – Tweed Byron Regional Council).

Poverty and inequality

2.11    'Poverty' and 'inequality' are distinct concepts although the terms are often used interchangeably. One submission has defined 'poverty' as a state where people lack access to the resources required to maintain a socially-defined 'acceptable' standard of life, whereas 'inequality' may be defined as a condition where people have fewer resources than others in society. It has been noted that poverty may increase even as inequality is reduced. This may occur, for example, if reductions in income inequality are brought about by increasing taxation to the point where it undermines economic incentives. Equally, it is possible for poverty to be reduced even as inequality widens. This has been the case in recent years in the United States, where child poverty has been progressively reduced since the 1980s even though tax cuts have almost certainly widened income differentials.[12]

2.12     The acceptance of poverty as a relative concept means that poverty and inequality are related. One study notes that:

Relative poverty can exist only where there is inequality, but this does not mean that poverty and inequality are the same. Not all forms of inequality will imply relative poverty, only those in which some individuals or groups fall well below the average. The eradication of relative poverty will require some reduction in inequality, but how much and at what cost will depend upon the nature and severity of the poverty problem.[13]

2.13    Many approaches to poverty are based on drawing a 'poverty line' which makes a distinction between the poor and the non-poor. Other researchers however tend to focus on questions of inequality rather than poverty per se, giving emphasis to the continuum of low incomes rather than to a single poverty line cut-off  point and focus on the consequences of low incomes for standards of material well-being and social participation. Inequality of income, alone, then assumes relatively less importance as a precondition for poverty, being combined with other dimensions of inequality, including inequality of access to stable employment, affordable housing, quality health care and government services more generally.[14]

2.14    Submissions to the inquiry noted that when discussing inequality it is important to consider all aspects influencing the various dimensions of inequality, including but not limited to income levels.[15] One submission noted that measures of inequality that focus on income can be misleading if they exclude other indicators such as access to education, health, government services and infrastructure.[16] Submissions also noted that traditional forms of inequality such as income inequality have been expanded in recent years to include 'new' inequalities such as the 'digital divide'.[17]

2.15    Submissions also emphasised that poverty in many cases is a precursor to other forms of disadvantage. A study for Jesuit Social Services identified a concentration of disadvantage using a composite measure of social disadvantage, including income levels, educational levels, unemployment, low birth weight, child maltreatment, childhood injuries, psychiatric admissions, mortality, crime and emergency relief. The study found most of these indices of disadvantage to be cumulative and mutually reinforcing. The broad concentration of disadvantage has been confirmed by a range of other studies.[18]

Measuring poverty

2.16    There are a number of ways in which poverty may be identified:

Each of these approaches is discussed below.

2.17    As noted earlier, estimates of poverty are usually estimates of relative poverty. Relative poverty is generally measured by drawing a 'poverty line' at an income level considered necessary to maintain an 'acceptable' standard of living. Professor Saunders pointed out that any measure of poverty adopted must be independent of the government of the day. He added:

...independent of government in the sense that we do not want to prejudge what we get out of it at the end of the day. If we are going to develop a measure that is going to endure, it has to last under today's government, tomorrow's government and the government in 10 or 20 years time, and we need to recognise that.[19]

Income-based poverty lines

2.18    Most Australian poverty research relies on income based poverty lines. These measures set a poverty line at some fraction of a statistically derived reference point derived from the overall distribution of income. They usually take the form of measures such as incomes below half the mean (average) income, or half the median income (that point where half earn more and half earn less than).

2.19    A common method of setting these poverty lines is to base the standard used on a measure of average community income. Generally some group in the population, such as a family comprising two adults and two children is chosen as the standard. For them, poverty is defined as having an income less than some fraction of the mean or median income. This is sometimes referred to as the 'headcount' index and can be used to measure the incidence of poverty among particular groups in the population, for example, the elderly, single adults, or couples with children. This approach was adopted by Professor Ronald Henderson when he developed his 'poverty line' which was originally set equal to the minimum wage plus child endowment in Melbourne in 1966.[20]

2.20    The advantage of income benchmarks are that they can be defined in terms of well-known statistical indicators and the poverty line which results is simple to update and understand. It also places the concept of poverty in the context of the distribution of income or resources within the community. But the choice of a particular proportion of average incomes to represent a 'low' income is necessarily an arbitrary decision.[21]

2.21    A major limitation of this measure of poverty is that it is sensitive to the level at which the poverty line is set. As poverty lines are typically set in income ranges where large proportions of social security recipients are clustered, small movements in the poverty line can result in large apparent increases or decreases in poverty. Another disadvantage of the headcount index is that it takes no account of the severity of poverty, that is, how far below the poverty line the poor actually are. As a result, government policies that raise the income of the very poorest will have no discernible impact on the headcount poverty rate if they do not raise the incomes of the poor above the poverty line.[22]

2.22    There remains a continuing debate as to whether such an approach should be based on the mean or median income level. The most frequently used measures include:

2.23    As noted above, many studies of relative poverty base their poverty line on some proportion of the median income of the population. This is calculated by ranking all income in descending order, establishing the income of the person in the middle of the income distribution ranking, and then defining the poverty line as some proportion of the income that they receive. As noted above, the proportion of median income chosen varies between studies.[23]

2.24    The main argument in favour of the median is that it is less sensitive to extreme (very low or very high) incomes. Against this, while mean income is more sensitive to extreme incomes it is dependent on the level of all incomes. A rise in the incomes of all those in the upper half of the income distribution would, for example, cause mean income but not median income to rise. If the poverty line were tied to median income, such a change would thus have no impact on the poverty line and hence on poverty, even though the relative position of those on low incomes would clearly have worsened.[24]

2.25    Some studies calculate poverty lines based on mean or average incomes. As many people receive fairly modest incomes but only a relatively few earn very high incomes, the mean income in the population is usually higher than the median. Basing the measurement of poverty on some proportion of the mean income will therefore usually produce a higher, more generous poverty line.

2.26    Defining poverty as some proportion of mean income is as arbitrary as using a median-based measure. Because the mean is sensitive to extreme values at either end of the income distribution, it follows that if a small number of people at the top of the distribution significantly increase their incomes this will inflate the mean and will therefore increase any poverty line based on it. This will in turn increase the number of people found to be 'in poverty', even though most people's living standards will not have changed.[25]

Poverty gap

2.27    In an attempt to overcome the disadvantages of the headcount approach an alternative poverty indicator – the poverty gap – is sometimes used. The poverty gap takes account of both the numbers below the poverty line and the depth of poverty. The poverty gap estimates the total gap between actual incomes and the poverty line for those who are in poverty. The poverty gap gives more weight to those whose financial situation is worst relative to their poverty line. The aggregate money value of the poverty gap indicates the minimum financial cost of raising all poor families to the poverty line.

2.28    As a measure, the poverty gap overcomes some of the shortcomings of the headcount ratio. It changes not only when people are moved from one side of the poverty line to the other, but also as a result of any change that increases the income of any poor family, where this is not offset by equal income reductions for other poor families.

2.29    The poverty gap is useful when analysing the impact of changes to income support payments on income poverty. The use of poverty gaps also makes poverty estimates less sensitive to small changes in the incomes of those close to the poverty line. This is because the poverty gap gives less weight to those closest to the poverty line and is thus less sensitive to changes in their income relative to the poverty line. A difficulty that the poverty gap does not overcome is that it is unaffected by redistributions of income among the poor.[26]

Budget standards

2.30    A second research method is the budget standards approach. This approach to defining poverty attempts to calculate 'objectively' what level of income people need to buy the goods and services they require to maintain an 'adequate' standard of living and construct a poverty line around this measure. The approach uses a combination of expert judgements and surveys of actual expenditures of people on low incomes to develop detailed 'minimum' household budgets. This approach prescribes a basket of goods and services that fulfils the 'necessities' of a standard family and then determines the average or minimum price of the basket. The basket of goods is usually defined in terms of food, clothing, shelter and other items needed for basic survival.

2.31    A number of studies have argued that this minimum income approach appears to provide a commonsense, simple and objective approach to poverty. One study notes that it 'appears to be divorced from personal values of either harshness or compassion. It seeks to describe poverty objectively as lack of the income needed to acquire the minimum necessities of life. Those who lack the necessities to sustain life are by definition poor'.[27]

2.32    Professor Saunders argued that the principal strength of this approach is its explicit attempt to derive the cost of meeting needs at a given standard of living across all areas of consumption. In addition, the fact that every item that enters the budget is identified precisely, as are the assumptions used for costing, means that the approach can be used to check what differences it makes if particular items are changed or omitted altogether. The approach also utilises prevailing social norms and standards to develop the budgets, thus building on existing knowledge about minimum standards where they exist. Finally, the budgets can be disaggregated to examine how much is needed to meet the standards in a limited range of areas, or to explore differences in the consumption levels of individuals within the family.[28]

2.33    An example of this approach is the project completed in 1998 by the Social Policy Research Centre (SPRC), funded by the former Department of Social Security, which produced a low cost 'indicative budget standard' for Australia. The work was based on focus groups of low income people and by examination of the actual expenditures of low and middle income households from ABS surveys.[29] The standard was derived by compiling a comprehensive list of items which most people buy or are deemed to need. The standard was calculated by pricing each item and calculating the total income which households of varying compositions require to pay for these items. ACOSS commented on the usefulness of this approach and that the next step was going to be a major study of living standards 'to look directly at what essentials people miss out on'. This project, however, was not undertaken which ACOSS argued was regretable as 'it leaves us to rely too heavily on income based poverty lines'.[30]

2.34    Several problems have been identified with this approach. A major problem is disagreement amongst researchers over what constitutes a 'minimum' amount of goods and services that would provide an acceptable standard of living. Specifically, it has been rarely possible for experts to specify minimum standards for necessities other than food. Allowances for other items are usually based on observation of the spending pattern of low income households – a procedure that is inconsistent and unsatisfactory. One study notes that 'the claim that the subsistence approach is scientific and pays minimum attention to value judgments is denied by the arbitrary judgments which have to be made in applying it'.[31] Professor Saunders also argued that this approach is very complex to operationalise and requires a large number of assumptions and judgements before the budgets can be specified and priced.[32]

2.35    Moreover, a subsistence standard has seldom been maintained over time – poverty lines based on this approach have generally become more generous in their poverty definitions in the light of changing circumstances.  It is also argued that this approach has little relevance in a society as diverse as Australia where living standards are generally far above the minimum required for merely physical survival.[33]

2.36    One study notes that a further problem with this approach is that there may be an 'upward bias' in the minimum income estimate because it does not allow sufficiently for the choices that are made between items that meet similar needs. The study suggests that the SPRC's calculation of a minimum low cost budget may be much higher than what is actually required to maintain a reasonable standard of living.[34]

Consensual approach

2.37    Another approach to measuring poverty is to base it on the opinion of the general population. Under this approach, sometimes referred to as the 'consensual approach', surveys are conducted where people are asked what they consider to be the minimum amount of income they need to 'make ends meet', and a poverty line is estimated around the point where the answers converge.

2.38    A number of studies have argued that these survey-based approaches to defining poverty lines appear logical given that the concept of relative poverty is defined with reference to prevailing social norms. If poverty is a standard of living below a socially accepted norm, it would appear to make sense to ask members of the public to define what that level should be.[35]

2.39    However, it has also been argued that these approaches have major difficulties. The answers to the questions asked have been found to be very sensitive to the wording used in the surveys. The studies also generally find that there is little consensus over what a minimum income might be – partly because people tend to be over-generous when defining minimum incomes for others which they will not have to pay for themselves, and partly because people's perceptions of what is 'necessary' for others tends to reflect the standard of living to which they themselves are accustomed. 'Poverty lines' based on this method tend as a result to be set at relatively 'generous' levels.[36] Given the diversity of opinion elicited in surveys the formulation of a 'consensual' poverty line is difficult to identify. Consequently, few studies have produced results using such poverty lines.[37]

Living standards studies

2.40    The weaknesses of income-based and budget standards methods have led many researchers to advocate direct measurement of poverty through surveys of people's access to various necessities of life.

2.41    While no Australian study has systematically attempted to measure deprivation (as distinct from living standards generally) across the community as a whole, many research studies commissioned by community agencies and Governments have examined aspects of deprivation within particular groups and communities. Some have used survey questionnaires, while many are anecdotal in nature. In Australia, a number of national studies relating to living standards in general (as distinct from deprivation) have been conducted since Professor Henderson's Melbourne survey.

The Australian Standards of Living Study (ASLS)

2.42    This study was conducted in the late 1980s by Richardson and Travers. It used a set of indicators to evaluate the living standards of a representative sample of all Australian households. The indicators ranged across economic resources, housing, employment, education, family and social relations, and recreation. The indicators were clustered into groups of common themes (e.g. social participation) and the population was grouped according to households' scores in regard to each cluster. The scores were then compared across a number of clusters to establish whether some people were consistently better or worse off.

2.43    The main advantages of this study were the breadth of indicators used, the ability to check whether the different dimensions of deprivation were related in a consistent way, and to compare them with the resources available to people in terms of income, assets, and family support.

2.44    However, the study had a number of drawbacks. The first is common to most general studies of living standards. Although many questions were asked to gauge people's general standard of living, few were developed specifically to measure deprivation.  The second drawback was the essentially arbitrary nature of the choice of indicators of living standards, and the development of indices of deprivation from the data. A third problem was that people were not asked whether they lacked an item through choice or lack of resources, a critical question for deprivation studies.

2.45    The study found that there were consistent relationships between deprivation in different dimensions of life, such as lacking material goods, poor health, and social participation. The study also found that there was a consistent relationship between income and disadvantage especially at the lowest income levels.[38]

The Australian Living Standards Study (ALSS)

2.46    This study was conducted by the Australian Institute of Family Studies in the early 1990s. It was also a comprehensive study of general living standards – using similar 'dimensions' of well-being to those adopted by Richardson and Travers.

2.47    One of the strengths of this study was the emphasis placed on how the respondents perceived the importance of each item, which somewhat reduced the reliance on the judgements of the researchers themselves. The researchers also asked people whether they lacked items through choice or a lack of resources.

2.48    The ALSS study shared a major drawback of the ASLS discussed above. It is a general living standards study with few indicators that go directly to deprivation. Moreover, it was also narrower in scope in that it covered only households with children in certain outer urban areas of Australia.

2.49    The ALSS study also found a consistent set of relationships between different dimensions of deprivation, and a clear relationship between deprivation and low income. The bottom 20 percent of households ranked by equivalent disposable income were found to consistently lack 'basic' resources and experience financial stress to a much greater extent than the top 80 per cent. This included such items as lacking money for school excursions, getting behind with bill payments and living in overcrowded housing.[39]

The ABS financial stress study

2.50    In the mid 1990s, the former Department of Social Security and the ABS collaborated to develop an official Australia-wide living standards survey, although public funding was not secured to complete this project. However, on the basis of the developmental work already under way, the ABS developed a module of 16 indicators of financial stress, which it added to its regular Household Expenditure Survey. The questions asked in the financial stress module included possession of certain household 'essentials', participation in social activities, the capacity to pay bills, raise money for an emergency, save, and action taken by people when they lacked resources to meet their needs (for example, pawning goods).

2.51    The main advantage of this survey is that it was incorporated into the ABS collection of major surveys of income and expenditure, providing a rich data source for comparing financial stress with the income and expenditure of different groups. The same survey (the HES) is often used by researchers to estimate the number of people with household incomes below income-based poverty lines. The ABS found that people's level of financial stress was inversely related to income, with much higher levels of stress being experienced among the bottom 20 per cent of households, although only a minority of them suffered moderate or higher stress.

2.52    ACOSS argued that the main weakness of the survey from the standpoint of poverty research is its small number of questions and narrow scope. Financial stress is only one indicator of deprivation, and it may mislead if used on its own. The survey did include indicators of a lack of basic household goods and restrictions on social life. However, these were so few in number as to heighten the general concern about the arbitrariness of the indicators used in deprivation studies. In such a small survey, the choice of indicators, and the way in which indices of deprivation are derived from them, is even more critical than in the larger living standards studies discussed above.[40]

2.53    A study by Bray used the financial stress study to measure the extent of deprivation across the community.[41] This study found consistent relationships between people's responses to certain clusters among the 16 indicators of financial stress. One cluster including items such as 'going without a meal' suggested a more severe form of deprivation, while another including such items as 'not having family and friends over once a month for a meal' suggested a less severe form of deprivation. On this basis, Bray developed two separate sets of indices of 'hardship' and 'missing out'. ACOSS argued that while these methodological judgements made sense conceptually, they were judgements based on limited data. The 'hardship' index relied on just four indicators, the lack of one of which constituted 'some hardship' while the lack of two or more constituted 'multiple hardship'. The results would be highly sensitive to the choice of indicators and the method used to develop the indices.[42]

Problems in measuring poverty

2.54    There are a number of difficulties in measuring poverty. These relate to the data sources used, questions related to the appropriate measure of income and other resources, the use of expenditure data, the income unit and time period used and the choice of equivalence scales. These issues are discussed below.

Data sources

2.55    Most poverty analysis in Australia is based on surveys of household income conducted by the Australian Bureau of Statistics (ABS). The ABS has conducted periodic income surveys since 1969. Detailed information relating to income from these surveys was first released for the 1982 Income and Housing Survey, allowing organisations to conduct analyses of income distribution and poverty. Since then, information has been released for the 1986 Income Distribution Survey, the 1990 Survey of Income and Housing Costs and Amenities and the Surveys of Income and Housing Costs conducted from 1994-95 to 2000-01, except for 1998-99.

2.56    Limitations in the scope and methodology of the income surveys affect the accuracy of poverty measurement. Although the methodology used for each survey has remained substantially the same over the years, the surveys have varied not only in frequency but also in scope, sample size and definitions used. In particular, there was a significant change in approach after 1990. The frequency of surveys was increased – they were conducted annually for four years and are currently conducted bi-annually – and the sample size has been halved. The definition of 'dependent children' has also changed over time.[43] Reporting on the 2000-01 income survey, the ABS stated that it had incorporated a range of methodological improvements in household income distribution and measurement designed to bring the data into line with international best research practice.[44]

Definition of income

2.57    The ABS income surveys measure gross cash income, which includes income from wages and salaries, self employment, government cash benefits, investments and other categories such as workers compensation, superannuation and royalties.

2.58    Some limitations of the ABS definition of income have been noted in several studies:

The measure of resources

2.59    A person's standard of living depends on a number of factors both tangible – such as access to income and/or wealth – and intangible such as the degree of satisfaction derived from work and other activities, or the extent of interaction with friends and/or the wider community. Such factors are difficult to measure directly so poverty measurement relies on some proxy or 'indicator of resources' that can provide a reasonable approximation of each person's standard of living.

2.60     Most poverty research examines the cash income received by families on the assumption that income provides a good guide to living standards. One study notes, however, that income is an indirect measure of poverty because it examines the resources on which living standards depend but does not directly measure the living standards actually experienced by families.[46]

2.61    Other studies have, however, emphasised the importance of income. The Henderson poverty inquiry emphasised the point that an adequate income is fundamental to a person's security, well-being and independence. It enables a person to provide for the basic essentials of life, permits freedom of choice and freedom to participate in activities of choice. It also contributes greatly to personal freedom and the extent of opportunities available.[47]

2.62    There are several measures of income that can be used to indicate a person's standard of living, including private income (such as wages and salaries, self-employment, interest, rent and dividends); gross income (private income plus government cash benefits); disposable income (gross income minus income tax); and final or social income (disposable income plus government non-cash benefits such as health or education services). Final income ought to provide the most comprehensive indication of a person's well-being but it is difficult to define and measure. Disposable income is easier to measure and like social income, reflects income after government interventions, thus giving a better indication of the income people have available to spend. Consequently, disposable income is the predominant indicator of resources used in poverty studies.

2.63    As noted above, the living standards of individuals and families are affected by a range of other factors in addition to their cash income, including the receipt of in-kind income such as fringe benefits and pensioner concessions and the provision of partially or fully subsidised government services such as Medicare. A recent study by the Smith Family, discussed further in chapter 3, analyses the impact of not only cash incomes and government transfer (social security) payments but also indirect taxes and indirect benefits, such as health, education, housing and welfare.[48] The ownership of assets may also be important. The economic wellbeing of two individuals on a similarly low income may be very different if one owns their home while the other is renting. This is an important issue when looking at poverty amongst the aged.

Discretionary income

2.64    Some commentators have argued that expenditure on items that are essential to living costs, such as costs related to housing, health, childcare, work-related costs and child support payments, should be subtracted from income to give a better indication of 'discretionary income'. Australian studies that have attempted to measure discretionary income have generally adjusted income for only housing costs. This is primarily because the income surveys only include housing costs and because the Henderson poverty inquiry set an important precedent by calculating poverty both before and after housing costs had been taken into account.

2.65    The case for using after-housing income relies on the fact that housing is such a large and essential expenditure for most families.[49] Families who are purchasing their home or renting privately will have much of their income 'locked up' in housing costs, reducing their other general consumption and associated standard of living. Their 'after-housing income' is significantly lower than their disposable income and so they are more likely to be in 'after-housing poverty'. By contrast, those families who own their own home or are in government housing typically have much lower housing costs. More of their income is available for general consumption than would otherwise be the case and so they are less likely to be in after-housing poverty. Another advantage of using the after-housing income approach is that it will vary with the different costs of housing in different regions, thereby reflecting one of the sources of regional variations in the cost of living.

2.66    An argument against the after-housing income approach is that, although having some form of housing is essential and not discretionary, there remains discretion in the quality of housing and the corresponding housing costs. A family that places a high priority on housing relative to other goods and services will spend more on housing and thus will have lower after-housing income. In effect, the after-housing measure distorts poverty measurement so that the families that place a high value on the quality of housing are more likely to be in poverty.[50]

Alternative measures of poverty

2.67    Given the difficulties involved in devising a comprehensive measure of a person's standard of living, several researchers have suggested that studies of poverty and deprivation should attempt to examine poverty using several indicators, try to incorporate qualitative studies of the non-monetary factors that affect poverty and incorporate life cycle and wealth studies that estimate the well-being of people over the course of their lifetime. In practice, such comprehensive studies are rarely conducted and poverty researchers generally content themselves with reporting on one facet of the complex mix of elements that influence the standard of living of members of society.[51] One study notes that a full understanding of poverty can come only from taking account of these varying approaches and attempting to blend them to give an overall picture of deprivation in society.[52]

2.68    Two main alternative approaches to income poverty measures have been suggested – these are the development of deprivation indicators and the use of expenditure data.

Deprivation indicators

2.69    A number of submissions argued that deprivation indicators need to be developed as alternatives to, or complementary measures to, income poverty statistics. Observing the level of deprivation experienced by those on low incomes is a way of identifying the income level that corresponds to poverty.[53]

2.70    Professor Saunders stated that deprivation measures provide a direct indicator of poverty status because they reveal that basic needs are not being met. This is in contrast with resource-based poverty indicators such as income, which are indirect in the sense that they imply that poverty is present, but do not actually confirm this through direct observation. Because deprivation reflects a situation of unmet need that is actually observed and thus experienced, it provides strong support for the claim that poverty exists. At the same time, it is possible to identify which forms of deprivation are regarded as unacceptable in the community by surveying public opinion, thus giving increased credibility to the measure.[54]

2.71    Deprivation indicators need to be broadly defined to include not only the material goods needed to meet minimum consumption standards, but also those activities that are broadly regarded as necessary to support participation in society. The list thus needs to include such items as being able to send a child on school excursions; having enough to afford to buy presents for family members on important occasions; having enough to pay household bills on time; and not having to rely on welfare agencies, or food or clothing banks to get by.

2.72    The deprivation approach has been employed in a number of recent European poverty studies, particularly in the United Kingdom. The UK Poverty and Social Exclusion Survey (PSE) provides a framework for defining poverty in terms of deprivation and exclusion indicators. The survey conducted in 1999 measured different dimensions of poverty and social exclusion in the lives of adults and children. The PSE poverty measure was developed in two stages. In a first survey, adults were asked which of a list of items and activities for children, and a second list for adults, they considered everyone should be able to afford in Britain today and which they should not have to go without. The list included food and clothing items; items and activities to allow children and adults to participate in society; and for children, items to assist educational development. In the second survey a different group of respondents were asked to identify items and activities that children and adults did not have because they could not be afforded. Results from both surveys were used to construct a poverty measure based on items or activities that children and adults had to go without to be defined as poor. For children and adults, a threshold of lacking two or more items or activities was established for this poverty measure.[55] Several submissions commented that the PSE was a good example of a recent survey along these lines.[56]

2.73    A variation on this approach is the Irish 'consistent poverty' approach, which combines an income-based approach and a deprivation approach. The concept of consistent poverty forms the basis of the Irish Government's National Anti-Poverty Strategy and defines a household as being in poverty if they have both low incomes and have a score of 1 or more on the deprivation scale. The Irish scheme is discussed further in chapter 18.

2.74    In Australia, the 1998-99 HES undertaken by the ABS collected data on a number of indicators of financial stress. These items may be considered a form of 'deprivation' measure in that they largely focus on items which a household may have gone without due to financial constraints. In the HES, information was collected on a number of items such as how households compared their standard of living with that of two years ago; if households are usually able to afford such things as a week's holiday away from home; have a night out once a fortnight or have friends over for a meal; and whether in the previous year, because of shortage of money, a household was unable to pay utility or insurance bills; sought financial help from family or friends or a community organisation; or went without a meal or heating.

2.75    Detailed analysis of the HES survey was reported in the study by Bray referred to earlier. This analysis identified that the set of financial stress questions used by the ABS fell into three discrete components. These are:

2.76    Submissions argued that a comprehensive set of deprivation indicators should be developed in Australia. ACOSS noted that they should be designed to measure the lack of key necessities required for subsistence as well as social and economic participation. Ideally, these should be clustered into a number of dimensions of living standards, or capacities to meet basic physical and social needs, for example, access to education and shelter. ACOSS argued that the appropriateness of these indicators in Australia should be tested by surveying the public as to whether they represent community opinion as to what constitutes a lack of necessities according to the definition of poverty used.[58]

2.77    Professor Saunders suggested that a step in this direction has already been taken by the ABS through the inclusion of a series of questions on financial stress and hardship in the latest 1998-99 HES (referred to above). He suggested that the ABS should expand its coverage in future HES and other household surveys or that a special social survey focusing on deprivation and its relationship to other indicators of living standards similar to the UK Poverty and Social Exclusion survey could be undertaken .[59]

2.78    FaCS suggested that deprivation approaches – although current data are relatively limited – as well as longitudinal analysis of household incomes, offer greater potential for policy and program development than income poverty measures.[60]

Use of expenditure data

2.79    Some submissions argued that expenditure data could also be used to supplement income data to measure living standards, given that reported incomes may be an unreliable indicator of a person's actual standard of living.

2.80    An expenditure poverty measure would rank people according to their expenditure, such that people with particularly low spending would fall into the category of 'expenditure poverty'. Evidence from ABS household expenditure surveys indicates that expenditure is consistently more unevenly distributed than income. This appears to be the case because people on high incomes devote some of their incomes to savings while people on low incomes, especially those temporarily on low incomes, may borrow to maintain their standard of living.[61]

2.81    A recent study for the Smith Family focused on the spending patterns of low income households, based on the 1998-99 HES. The study found that:

2.82    The use of expenditure data, however, has a number of problems. The most notable is that generally expenditure on consumption rather than consumption itself is being measured. Furthermore, it may be possible to sustain a standard of living during periods of low income by running down past savings or incurring debts, but this situation will not be sustainable. Low consumption may thus conceal rather than reveal the existence of poverty.[63] Professor Saunders suggested that it may be fruitful to consider income and expenditure not as alternative indicators of poverty, but rather as complementary.[64]

The income unit

2.83    While poverty calculations rank people according to income and then identify those who fall below a predetermined poverty line it is also important to establish whose income they should be ranked by. The income unit refers to the group within which income is assumed to be shared. The assumption widely accepted in academic research is that in most families the income earned by each member of the family is shared.

2.84    Possible income units include the individual, the nuclear family, the extended family or the household. The income unit used in poverty studies can make a major difference to poverty estimates. Generally, the broader the definition of the income unit, the lower the level of poverty is likely to be. The most common income unit in Australian poverty research is the 'ABS income unit' The ABS defines an 'income unit' as a couple (married or de facto) with dependent children; a couple without dependent children; a sole parent; or a single person. Income is assumed to be shared by one person or a group of related persons within a household. Income sharing is assumed to take place within married or de facto couples, and between parents and dependent children.

2.85    No income unit can capture the 'income-sharing group' precisely. One problem common to the ABS income unit, the extended family and the household is the appropriate treatment of dependent children. As they get older, children gradually gain greater financial independence but the exact point at which they are predominantly self-sufficient will vary greatly between families. This suggests that each of the potential income units may be too broad because they include people where little income sharing remains.

2.86    Some commentators have argued that common definitions of the income unit are too narrow. For example, young people living away from home, and therefore treated as separate income units, may well receive substantial financial support from their parents and thus not really be an independent unit. Similarly, it has been argued that different cultural attitudes towards income-sharing, particularly among Indigenous communities, often mean that income is shared much more widely amongst these communities than the ABS income unit or even the household.

2.87    FaCS indicated that with most income analysis it prefers the household as the basis of analysis. One reason for this is the structure of income support payments, for example living at home rates, which acknowledge the presence of others in the household. Further, even where income pooling might be partial, people living in the house generally gain some benefits from the expenditure and assets of others, ranging from the use of household appliances to sharing transport.[65]

2.88    There has been little research undertaken in Australia that examines the most appropriate definition of the income unit for poverty analysis as such a study requires information on the intra-household allocation of resources – for which there is little data.[66]

Time period

2.89    As household incomes fluctuate over time, the period over which the income is received can influence results. The time period used is frequently dictated by the available data. Time periods that are frequently used within sample surveys include the week, the month, and the year. Generally, longer time periods result in more equal income distributions. The ABS income surveys include both 'current weekly income' (that is, income in the week immediately preceding the relevant survey) and annual income from the previous financial year.[67]

2.90    FaCS noted that Australian income distribution analysis is unusual because of the general reliance placed on using current income rather than annual income. The main reason for this is the general view that current income is more accurately reported, especially in relation to income support payments. The consequence of this is that poverty estimates using current income may indicate greater income inequality whereas the use of annual income would smooth out fluctuations in household incomes.[68]

Equivalence scales

2.91    Equivalence scales attempt to account for the differences in the size and composition of different income units – they show the amount of income required by families of different size and composition. Larger income units need a greater income than a small income unit to attain the same standard of living. The composition of income units also needs to be accounted for as some members of the income unit create greater 'costs' than others. For example, it is likely that an income unit needs a greater income to support older children than younger ones. Equivalence scales show how much more income a couple requires to reach the same standard of living as a single person. For instance, if an equivalence scale gives a single adult a value of 1 and a couple with no children a value of 1.7, then the assumption is that the couple requires 70 per cent more income than the single person to reach a comparable standard of living. The value is less than 2 because of assumed economies of scale, such as the sharing of housing costs. The equivalence scale used can have an important impact on the apparent degree of poverty and on the characteristics of those deemed to be in poverty, thus an equivalence scale that ascribes high costs to children will tend to place more families with children in poverty.

2.92    There is no agreement in Australia about what is the most appropriate equivalence scale to use. The Henderson equivalence scale has been used in many poverty studies in Australia. The detailed Henderson scales take account of the needs of families that vary with the age, gender, labour force status and number of adults and children within a family.

2.93    The trend in international comparative studies has been towards simpler equivalence scales that only vary with the number of people – or sometimes the number of adults and children – within a family. The equivalence scale adopted by the EC Eurostat poverty studies allows the assumed needs of a family to vary in line with the number of adults and children – with the first adult being given a value of 1, second and subsequent adults a value of 0.5 and children a value of 0.3 each. One disadvantage of this equivalence scale is that it does not allow for economies of scale, so that, for example, four children cost four times as much as one child.[69]

Updating poverty lines

2.94    Once a poverty line has been established it is important to determine how it should be updated over time, especially to determine trends in poverty over time and for assessing the adequacy of income support payments. Poverty lines can be updated according to movements in prices or movements in incomes. Updating by movements in prices adjusts the poverty line for changes in the cost of living whereas updating for movements in incomes adjusts for changes in the standard of living. The consumer price index (CPI) is typically used to update poverty lines in accordance with movements in prices.

2.95    There are several ways to update poverty lines in accordance with movements in incomes. Studies in Australia have variously used average weekly earnings, household disposable income per capita and movements in average or median equivalent disposable incomes. While there is debate about which of these methods is most appropriate, updating in accordance with movements in incomes has generally been preferred to movements in prices in Australia. This is related to the preference for relative poverty lines.[70]

2.96    The poverty line used by the Henderson poverty inquiry to estimate poverty in 1972-73 was set at 56.5 per cent of seasonally adjusted AWE in 1972-73. The Henderson poverty line was designed to be compared with after-tax or disposable income but the poverty line was updated (until the 1980s) according to movements in before-tax seasonally adjusted AWE. In response to criticisms of this method of updating the poverty line, since 1981 the Henderson poverty line has been updated by increases in seasonally adjusted household disposable income per capita. It had been argued that movements in disposable income per capita provides a better measure of changes in community living standards than movements in pre-tax average weekly earnings.[71]

2.97    This measure has in turn been criticised by some researchers because it has led to a substantial increase in the value of the poverty line, well in excess of increases in the CPI or AWE – 'this means that the Henderson poverty line has become less and less 'austere' as time has gone on, for it has been inflating at twice the rate of the CPI'.[72] Other commentators have suggested, however, that if it is accepted that poverty is relative, then it is logically appropriate that the poverty line should increase in real terms when real community incomes are growing and that seasonally adjusted household disposable income per capita or a similar index is the appropriate measure to use in these circumstances.[73]


2.98    Poverty is difficult both to define and to measure. The varying concepts of 'poverty' employed, data limitations and different methodologies all impose limitations on the degree to which poverty can be quantified. Poverty relates to individual standards of living which are influenced by a range of tangible factors – such as access to income and other financial resources – and intangible factors, such as social and family ties, which are largely unquantifiable but which can influence a person's overall standard of living. Despite these limitations, research into poverty and the development of measures of poverty are important in establishing the nature and extent of poverty in the community and in the development of appropriate policy responses.

2.99    Further research in the area of poverty can only lead to a broader understanding of its causes and effects and lead to improved ways to address the problem. The St Vincent de Paul Society noted that 'regrettably, the polemical debate over poverty lines has distracted rational discussion on solutions to a known problem'.[74] The Committee concurs with this sentiment and firmly believes that protracted debate over various poverty measures or the suitability of different poverty lines should not overshadow a thorough and ongoing discussion of solutions to a very serious national problem.

Navigation: Previous Page | Contents | Next Page