Position paper on poverty measurement

POVERTY MEASUREMENT A POVERTY AND INEQUALITY PARTNERSHIP POSITION PAPER MAY 2025 ACOSS AND UNSW SYDNEY Acknowledgements Our thanks to all those consulted for this report, including Poverty and Inequality Partnership supporters, ACOSS members, and poverty researchers. This report was written by Dr Peter Davidson at needtoknow consulting, Dr Yuvisthi Naidoo and Associate Professor Bruce Bradbury at the Social Policy Research Centre, UNSW Sydney. © Australian Council of Social Service (ACOSS) and UNSW Sydney 2025 Poverty measurement: A poverty and inequality partnership position paper is published by the Australian Council of Social Service, in partnership with UNSW Sydney. Locked Bag 4777 Strawberry Hills NSW 2012 Australia Email: [email protected] This report is an output of the Poverty and Inequality Partnership, between ACOSS and UNSW Sydney. Find out more at http://povertyandinequality.acoss.org.au ISBN: 978 085871 111 2 DOI: https://doi.org/10.5281/zenodo.13534526 Licence: https://creativecommons.org/licences/by/4.0/ Suggested citation: Davidson, P, Naidoo, Y., and Bradbury, B. (2025), Poverty measurement: A Poverty and Inequality Partnership position paper , Sydney: Australian Council of Social Service (ACOSS) and UNSW Sydney. Photos by istockphoto.com. All images are representative only. Poverty measurement: A poverty and inequality partnership position paper 2 ACOSS Partners DAVID MORAWETZ SOCIAL JUSTICE FUND JOHN MITCHELL 3 Glossary ACOSS Australian Council of Social Service Budget standards Minimum household budgets developed to calculate the cost of essential goods and services needed to achieve a socially acceptable living standard. Disposable income Household income from all sources (e.g. wages, benefits) minus income tax, typically used as the basis for poverty measurement. Equivalence scale A method used to adjust household income based on size and composition, ensuring comparisons of living standards across different household types. Commonly used scale: the modified OECD scale. Henderson poverty lines Australian poverty threshold derived from the 1975 National Poverty Inquiry, originally linked to the basic wage and later to average earnings. HILDA survey Household, Income and Labour Dynamics in Australia - a longitudinal study. Imputed rent An estimate of the value of housing services that homeowners receive from living in their own homes. Used to reflect housing resources in poverty research. Material deprivation A form of direct poverty measurement where individuals lack key goods or services deemed essential by the community, due to financial constraints. OECD equivalence scale A standard equivalence scale where the first adult is assigned a value of 1, additional adults 0.5, and each child under 15 a value of 0.3. Multidimensional Poverty Index (MPI) A composite measure of poverty that incorporates various dimensions such as education, health, living standards. Person Level Integrated Data Asset (PLIDA) An integrated data resource that links tax, social security and other administrative datasets for research. Multidimensional Poverty Index (MPI) A composite measure of poverty that incorporates various dimensions such as education, health, living standards. Social Transfers in Kind (STIK) Non-cash government benefits such as subsidised healthcare or housing. Survey of Income and Housing (SIH) Regular survey by the Australian Bureau of Statistics (ABS) providing detailed information on income, housing and wealth. Poverty measurement: A poverty and inequality partnership position paper 4 Contents Foreword 7 Executive summary 8 Detailed recommentations for income-based poverty measures 10 1. Background: Purpose and development of poverty research in Australia and internationally 14 1.1 The purposes of poverty measurement 14 1.2 The development of poverty research 16 1.3 The National Poverty Inquiry and ‘Henderson Poverty Lines’ 16 1.4 Increasing diversity of research into poverty and disadvantage (1980s-2020s) 16 1.5 Recent poverty research in Australia 16 1.6 Official poverty measures in other wealthy nations 17 2. Defining poverty and its relationship with wider research on social and economic disadvantage 18 2.1 A definition of poverty 18 2.2 The relationship between income-based poverty measures and direct measures of inadequate living standards 18 2.3 Bringing the voice of people affected into poverty research 21 3. Measuring income-based poverty 24 3.1 Key requirements for accurate and impactful poverty measurement 24 3.2 Income: definition and scope 24 Recommendations: Definition of income 28 3.3 Adjustments to better capture household spending power 28 Recommendations: Adjustments to capture major variations in living costs 30 3.4 Poverty thresholds 31 Recommendations: Poverty thresholds 37 3.5 Data gaps and reliability 38 Recommendations: Data availability and reliability 39 5 4. Outputs of poverty research 41 Recommendations: Outputs of poverty research 43 5. A transparent and robust process to develop official poverty measures 44 Recommendations: An open and robust process to develop official poverty measures 44 Attachment 1: Recent income-based poverty measures used in Australian research 46 Attachment 2: Poverty measures used in other wealthy nations 48 Endnotes 50 Poverty measurement: A poverty and inequality partnership position paper 6 This Poverty measurement: A Poverty and Inequality Partnership position paper is the 25th paper from the Poverty and Inequality Partnership between ACOSS and UNSW Sydney. This Partnership was established in 2017 to address poverty and inequality in Australia through research and advocacy. We have consistently reported on the nature and extent of poverty through our Poverty in Australia series, using the 50% of median poverty line (as used by the OECD) in the absence of an official Australian poverty measure. This paper seeks to inform the development of official poverty measures in Australia. It recommends the adoption of a dual approach, which includes income-based metrics complemented by direct or multidimensional metrics. The paper assesses a range of income-based poverty metrics and makes recommendations about the substance and process to address this critical gap in national progress indicators. The Poverty and Inequality Partnership relies on the support of non-government organisations including 54 reasons (part of Save the Children Australia Group), ARACY, cohealth (a Victorian community health service), Foodbank Australia, Jesuit Social Services, Life Without Barriers, Mission Australia, SSI, and The Smith Family. Additionally, it is supported by the Social Justice Fund, part of the Australian Communities Foundation, and John Mitchell. We would like to thank UNSW’s President and Vice-Chancellor, Professor Attila Brungs; Vice-President Societal Impact, Equity and Engagement, Professor Verity Firth; Emeritus Professor Eileen Baldry; and the ACOSS Board for backing this Partnership and its important work. Foreword Cassandra Goldie ACOSS Carla Treloar Social Policy Research Centre, UNSW 7 Executive summary Purpose of this report The purposes of this position paper on poverty measurement are to: • reflect the growing consensus among researchers and advocates on incomebased poverty measures specifically, and the need for a dual approach to measuring poverty comprised of income-based and direct (including multidimensional) measures, and to support further dialogue; • inform the future methodology of the Poverty and Inequality Partnership’s research using income-based poverty measures; and • inform the development of official poverty measures, closing a major gap in Australian social research and policy. Origins and scope of this position paper In late 2024 the Poverty and Inequality Partnership between ACOSS and UNSW Sydney undertook consultations with our research partners, ACOSS members and poverty researchers in other Australian research institutions to help build consensus on income-based poverty measures. This included a discussion paper and two online meetings. This paper is an outcome of those discussions. At the same time, the Brotherhood of St Laurence and the Melbourne Institute undertook consultations on a Multidimensional Poverty Index for Australia based on the work of Alkire and associates at the University of Oxford.1 These included a discussion paper, a seminar with Professor Alkire, and a conference in Melbourne in November 2024. This paper focusses on income-based poverty measures. After briefly examining the history of poverty measurement in Australia and teasing out the relationship between income-based and direct or multidimensional measures and how research can be informed by lived experience, it makes a series of recommendations to inform official income-based poverty measures and our own ‘Poverty in Australia’ research series.2 The problem – unlike most nations, Australia lacks official measures of poverty Unlike most other nations, Australia lacks an official poverty measure. It has been left to academic researchers and community organisations (including the Poverty and Inequality Partnership) to carry the torch for poverty research and hold governments and other institutions to account for the scale and persistence of poverty in Australia • Australia has adopted the United Nations Sustainable Development Goals , which require us to reduce poverty over time by reference to an official national definition. • However, official measures of wellbeing such as the Measuring What Matters framework are silent on poverty. Poverty measurement: A poverty and inequality partnership position paper 8 • The government’s Economic Inclusion Advisory Committee (EIAC) was tasked with providing advice on economic inclusion including policy settings, systems and structures, and the adequacy, effectiveness and sustainability of income support payments , yet poverty reduction was not included in its Terms of Reference.3 • EIAC has recommended that official measures of poverty and targets to reduce it be developed by government. Defining poverty People are in poverty when their income and resources are inadequate to attain a socially acceptable living standard. 4 It follows that poverty research can be divided into two strands: measuring economic resources and unacceptable living standards . The familiar income-based poverty measures such as ‘poverty lines’ are measures of economic resources – the minimum needed to avoid poverty. As such they measure living standards indirectly. Other measures aim to assess the adequacy of living standards directly. They include uni-dimensional measures, such as financial stress; and multidimensional measures, such as multiple deprivation and multidimensional poverty indicators that assess living standards across different domains of wellbeing - for example nutrition, housing, health, social isolation, and education. These two strands of poverty research are mutually reinforcing: • Income-based measures identify people likely to face poverty due to a lack of economic resources and set benchmarks for the adequacy of minimum incomes. • Direct measures of inadequate living standards can be used to validate income-based measures, capture the impacts of poverty on people across a range of dimensions, and inform the development of policies to ease poverty that extend beyond income support, such as education, health and housing policies. Attachment 1 lists examples of recent income-based poverty research in Australia. Attachment 2 lists several official income-based poverty measures adopted by other wealthy nations and influential international institutions such as the OECD. 9 1.2 The development of poverty research Poverty research has a long international tradition, dating back (in the United Kingdom at least) to the industrial revolution of the 19th century. Over time, poverty research has developed from a core focus on income or consumptionbased measures (‘poverty lines’) to research that maps and then tracks the effects of poverty across dimensions of life.8 1.3 The National Poverty Inquiry and ‘Henderson Poverty Lines’ In Australia the extent of poverty has only been rigorously measured by social scientists over the past 60 years, beginning with Professor Henderson’s pathbreaking survey of poverty in Melbourne in 1966. He was later commissioned by the Australian government to lead a national poverty inquiry, which reported in 1975. The centrepiece of that report was research into the extent of poverty across Australia and the profile of people affected, based on set of income thresholds later called ‘Henderson Poverty Lines’.9 Despite the official status of the report from the Henerson Poverty Inquiry and the rigour of the analysis undertaken, the then Australian government did not endorse the Henderson Poverty Lines nor any other metric as an ‘official’ measure of poverty. 1.4 Increasing diversity of research into poverty and disadvantage (1980s-2020s) During the 1980s and subsequently, poverty research was facilitated by improvements in income and other data, microsimulation models, and the emergence of a diversity of academic institutes undertaking research into poverty, income inequality and the impact of public policies upon them: • In Australia new measures of poverty were developed and debated, including the ‘50% of median income’ benchmark used by the OECD, budget standards, and ‘direct’ measures of material deprivation and financial stress.10 • These measures were often developed in collaboration with (or in response to) the work of researchers in other nations, especially those with which Australia usually compares itself (such as the United Kingdom, New Zealand Canada and Ireland, together with the European Union). • From a wider international perspective, United Nations agencies have promoted measures of poverty and wellbeing that can be compared across nations with different levels of wealth and income. These are often multidimensional – cutting across different domains such as income, health, education and housing – for example the United Nations Sustainable Development Goals, to which Australia is a signatory.11 1.5 Recent poverty research in Australia There is now a great deal of consistency in the use of ‘income-based poverty’ measures as defined below, and broad agreement that these should be complemented by more direct measures of deprivation including multidimensional approaches. Poverty measurement: A poverty and inequality partnership position paper 16 • Attachment 1 lists examples of recent income-based poverty research in Australia. Use of the 50% and 60% of median income poverty lines and the ‘modified OECD equivalence scale’ are commonplace, along with ‘before and after-housing’ poverty thresholds (these concepts are discussed later). 1.6 Official poverty measures in other wealthy nations Over time, many wealthy nations and influential international institutions have adopted official poverty measures, leaving Australia a laggard in this field: • Attachment 2 lists some of these measures, which include the 50% of median income poverty lines (used by the OECD) and the 60% of median income poverty line (used by the European Commission) as well as multidimensional measures.12 Governments in several comparable countries have undertaken regular official studies of the incidence of poverty and set poverty reduction targets to inform and steer public policy. 17 2.1 A definition of poverty The first and vital step is to define what we are measuring: People are in poverty when their economic resources are inadequate to attain a socially acceptable living standard. 13 While (as discussed below) it is an imperfect measure of economic resources, income sits at the centre of most poverty research in Australia. As the report of the Commission of Inquiry into Poverty argued: Income is fundamental to a person’s security, wellbeing and independence. It enables him [sic] to provide housing, education, food, transport and other essentials …. freedom of choice and freedom to participate in activities of his choice. 14 To determine who lives in poverty we need to decide how to measure economic resources and set the resource threshold(s) consistent with an acceptable living standard. Different theoretical frameworks have been used to inform these decisions in poverty research: • In a utilitarian framework, the metric used to assess a minimum acceptable living standard is people’s consumption possibilities (the set of ‘essential’ goods and services people can consume to meet their basic needs and participate in society). • In Sen’s capabilities framework two metrics are used: the resources required to give people the ‘capabilities’ to lead a life they value (noting that capabilities are not directly observable) and the ‘functionings’ or outcomes achieved in different domains of life.15 While the underlying theoretical frameworks are different, in practice there is much convergence between the measures of economic resources used in these two frameworks. 2.2 The relationship between income-based poverty measures and direct measures of inadequate living standards We propose a dual track poverty measurement framework, outlined below, with two streams of poverty research based on the above definition of poverty as having insufficient resources to attain an acceptable living standard: 1. Income-based poverty measures, and 2. Direct measures of inadequate living standards.16 2. Defining poverty and its relationship with wider research on social and economic disadvantage Poverty measurement: A poverty and inequality partnership position paper 18 To distinguish between the above two streams of poverty research, • income-based measures are indirect measures of poverty, focussing on the primary input to poverty (lack of economic resources); • direct measures focus on the outcomes and experience of poverty (e.g. inadequate diet, social isolation, a struggle to make ends meet) and factors beyond income that contribute to poverty (such as limited formal education or lack of paid employment). Income-based poverty measures This stream of poverty research measures people’s spending power (their ability to buy essential goods and services) including: • the adequacy of their incomes (typically using poverty lines );17 • supplemented by broader measures of economic resources such as wealth, access to affordable housing and/or other forms of in-kind income; and • validated by ‘budget standards’ research that sets minimum household budgets to achieve a socially acceptable living standard.18 They are the main focus of this paper and are discussed in detail later. Direct measures of inadequate living standards The three main strands of this research are: 1. Measures of multiple disadvantage across a number of domains of living standards such as material deprivation or a Multidimensional Poverty Index In material deprivation research , people are asked whether a list of goods and services (such as a decent and secure home) are essential, whether they have 1. Measures of inadequate economic resources Income-based ‘poverty lines’, taking account of other economic resources (e.g. housing) Informed and validated by: Minimum standards for consumption of essentials (including Budget Standards) 2. Direct measures of an unacceptably low living standard Measures of the outcomes of inadequate economic resources across different domains of wellbeing (including deprivation of essentials and multidimensional poverty measures) All guided and informed by: Lived experience of people directly affected Proposed dual track poverty measurement framework 19 them, and if not whether this is because they cannot afford them (in which case they are ‘deprived’ of an ‘essential’). A measure of multiple deprivation is calculated based on the number of households deprived of a threshold number of essentials.19 The Multidimensional Poverty Index developed by Professor Alkire and colleagues at the University of Oxford sets thresholds for disadvantage within each of several domains of wellbeing (such as employment, health and education). They apply a set of statistical rules to combine measures of the extent of disadvantage across multiple domains into a single index.20 Key benefits of multidimensional measures include that they enable researchers and policymakers to capture the impact of poverty across different domains of wellbeing and to better understand how disadvantages in different domains interact: • for example, the relationship between low education, labour market marginalisation and poverty, or between poverty and poor health. One risk is that material deprivation, especially inadequate income, may be downplayed when measuring multidimensional disadvantage, and in policy responses: • This was a widespread concern among poverty researchers and public policy advocates when the former UK Government abandoned income-based child poverty targets and replaced them with a different set of targets focussed on employment participation, education, and parental responsibility.21 Multidimensional approaches have generally been developed as complements rather than substitutes for income-based poverty measures: • For example, Alkire and colleagues recommend that national income poverty measures be developed alongside a wider multidimensional index.22 2. Financial stress Financial stress is self-reported and refers to a lack of resources to meet current needs. There is some overlap between this and the broader material deprivation research described above (for example, the inability raise a certain amount of money in an emergency is a variable that may be used in both, however, the latter seeks consensus on items being considered essential).23 3. Reporting people’s experience of living with low economic resources. This research typically uses focus groups or surveys of people considered to be at risk of poverty, to better understand how living with very limited resources is experienced.24 The pros and cons of direct and income-based poverty measures Direct measures of poverty have important advantages over income-based poverty lines: • Ultimately, it is the outcomes of poverty that matter for people affected and for public policy. • By capturing people’s living standards across a number of domains (e.g. access to employment, affordable housing and health care) they add depth Poverty measurement: A poverty and inequality partnership position paper 20 and nuance to poverty measurement that can be linked to specific policy actions. • They position the lived experience of people directly impacted as central to discussions on poverty measurement. Direct measures of poverty, however, have their own limitations: • They require researchers to juggle a wider and more complex range of variables. • To chart trends in poverty using direct measures such as multiple deprivation, a diverse set of indicators must be converted into an index. While this may be more meaningful in theory than a poverty line based on income alone, this requires careful expert judgement – especially over which indicators to measure, what thresholds to set, and whether and how to convert this to meaningful income measures to facilitate public understanding. • Direct measures require regular adjustment to account for changes in community living standards and expectations (such as growing reliance on information technology to meet basic needs). These adjustments should be made regularly and gradually, to avoid conservative bias and abrupt changes in poverty (as measured) that reflect changes in methodology rather than real world experience. • Self-assessed poverty is subject to measurement error due, for example, to differences in cultural framing of poverty and the stigma that is attached to identifying as ‘poor’. As in most social research, no single measure gives us the complete picture of who lives in poverty, what causes it, how it is experienced, and the individual and social impacts on people’s wellbeing. We need to triangulate between income-based and direct measures. • One option is to use direct measures to validate income-based poverty lines, for example by identifying income thresholds above which the incidence of material deprivation rises sharply.25 • Another option is to regularly conduct research using both income-based and direct approaches and compare the results. To the extent that the same groups are identified as at risk of poverty using both measures, we can be more confident that they live in poverty.26 2.3 Bringing the voice of people affected into poverty research To comprehensively research poverty we must understand how people experience it. The most comprehensive poverty research has been informed by feedback from people directly affected about how they came to live in poverty, its effects on their lives, and what would lift them out of it.27 In recent years, poverty researchers have sought guidance from people directly affected to set research goals and questions and develop the methodology used. Poverty is disempowering. It excludes people from the economic, social and political mainstream. Researchers who have not experienced poverty directly 21 or worked closely with people who have, cannot assume they understand its meaning or impact. It is especially important for poverty researchers to engage with and hear the voices of people from communities and groups that are also marginalised in other ways and/or are likely to understand poverty from a different value system, including: • First Nations communities (discussed below); • Children and young people, who experience poverty differently to other family members, even though decisions to classify them as ‘living in poverty’ are generally made on the basis of their family’s circumstances. Since poverty can have lifelong impacts on children, researchers should understand how they experience it; • People with disabilities; and • People with diverse cultural and linguistic backgrounds. 28 Poverty in First Nations communities To define and accurately measure poverty in First Nations communities, researchers first need to know how poverty is understood and experienced by people in those communities. For example, while they are well aware of their relative lack of material resources and essential services, ‘deficit framing’ of First Nations communities is problematic.29 Researchers should collaborate with First Nations communities and experts drawn from those communities when determining the purpose, design and conduct of research including the dissemination of results. The methods used and interactions with First Nations people when conducting poverty research should be respectful and culturally safe. The research and any associated public policies should contribute to the betterment of peoples’ lives as defined by the communities themselves. By taking this approach, researchers have found that poverty is understood and experienced differently in First Nations communities, for example: • priorities for what constitutes a ‘good life’, including in many cases living on country (where essentials are often more costly) and giving priority to sustaining cultural and ceremonial practices; • kinship structures and obligations, which have implications for the way in which resources are shared. • Housing costs per person may be relatively low, but homes are often of poor quality and overcrowded.30 This suggests that poverty should be defined and measured differently in First Nations communities, while retaining sufficient common ground to facilitate comparisons with living standards in the broader community (for example to measure progress in Closing the Gap targets).31 • These measures can only be developed by working with First Nations communities and experts from the outset to determine the purpose of the research, how poverty is defined and measured, and the policy implications of the research. Poverty measurement: A poverty and inequality partnership position paper 22 23 The following discussion focuses on methodology for income-based measures of poverty , with some reference to the direct measures of poverty identified earlier. 3.1 Key requirements for accurate and impactful poverty measurement The following criteria are proposed for setting income-based poverty measures: 1. Validity and accuracy : especially the capacity to identify the proportion of people living in poverty, who they are, and the depth or severity of poverty among those affected; 2. Timeliness: to track trends in poverty and their possible causes; 3. Simplicity and ease of public understanding; 4. Consistency of measurement: to facilitate tracking of long-term trends and avoid confusion among followers of poverty research; 5. Policy relevance : to inform analysis of the causes of poverty and policy solutions; 6. Legitimacy : consistent with the experience of poverty among people affected and community views of poverty. Each of these criteria is important, both to improve understanding of poverty and to develop policies to reduce it. It is unlikely that a single measure will meet all of them. 3.2 Income: definition and scope Definition of income In a market-based economy such as Australia’s, income is the primary determinant of living standards for most people since it is the main determinant of their spending power:32 • Income rather than consumption is mainly used in poverty measurement in wealthy countries since consumption is affected by people’s choices – for example they may choose to reduce current consumption to set aside income for retirement or to save to buy a home in a more desirable neighbourhood. For its Survey of Income and Housing (SIH) , the ABS defines income as: all current receipts, whether monetary or in kind, that are received by the household or by individual members of the household, and which are available for, or intended to support, current consumption. 33 3. Measuring income-based poverty Poverty measurement: A poverty and inequality partnership position paper 24 The components of cash income generally included in ABS income surveys and most poverty research include private income (primarily earnings and income from investments) together with social security payments minus income tax. • This is generally defined as disposable income. Disposable income as defined in ABS income surveys is narrower than definitions of income commonly used in economic theory, which also includes increases in wealth (e.g. capital gains on the value of people’s homes) – the role of wealth in poverty measurement is discussed below.34 Income according to the above definition is not limited to cash. For example, some goods and services are available to people with low incomes free or at low cost through government programs. Whether the ‘in kind’ value of these goods and services should be considered in poverty measurement is discussed later. The period over which income is measured The period over which income should be measured depends on the purpose of the research: • The focus of income-based poverty research is usually people’s current living standards, generally understood to refer to weekly or annual income.35 • Annual income is a more comprehensive and consistent measure than weekly income, since weekly incomes often vary substantially especially for people on the margins of the labour market and those with modest investment incomes. • However, annual measures have their own problems, including data availability. For example, the ABS no longer includes estimates of annual income in its Survey of Income and Housing; and survey respondents may not accurately recall their incomes over longer periods. • The persistence of poverty over a number of years is an important topic for poverty research. For such analysis a longer reference period may be appropriate (beyond simply estimating whether people are experiencing poverty for a certain number of years in succession) and more components of wealth (e.g. the stock of superannuation savings) may then be relevant (as discussed below). Wealth Income and wealth are separate but related. Wealth is a ‘stock’ measure (the assets people hold at a point in time) and income is a ‘flow’ measure: • Income is equal to consumption in a given period (such as a week or a year) plus the change in wealth from the beginning to the end of that period.36 Although most drawdowns of wealth are excluded from disposable income as measured in the ABS income survey (though regular superannuation payments are included), the wealth available to people to spend also affects their current spending power: • Depending on the form that wealth takes, people can draw down their stock of wealth to buy more goods and services.37 25 • A poverty threshold based on a basket of commodities, such as those developed for budget standards research in Australia (discussed later); or • A relative poverty line (such as 50% of median income) that is ‘frozen’ in real terms over a period – an uneasy compromise between ‘relative’ and ‘absolute’ approaches. Both approaches still require checking from time to time on whether the poverty threshold is consistent with current understandings of a minimum acceptable living standard: • A poverty threshold that is frozen in real terms for many years will likely fail to take account of changing community standards, for example widespread uptake of mobile phones as an essential communication device. • It is generally accepted that the basket of commodities used for budget standards research should be updated approximately every five years. This raises the possibility that the rate of poverty as measured using budget standards benchmarks may change abruptly as the budgets are updated, complicating any assessment of trends in poverty. 50% and 60% of median income The most commonly used poverty thresholds in Australia (see Attachment 1) are based on a percentage of household median equivalised disposable income, where: • The poverty line for a single adult living alone is set at 50% of the median income, consistent with the primary income-based poverty measure used by the OECD (there are different poverty lines for different sized households, as discussed later). • In some cases (including in Poverty and Inequality Partnership poverty research), the higher European Commission threshold of 60% of median income is used in conjunction with the 50% threshold for sensitivity testing purposes.54 The use of median income (as distinct from other points in the income distribution) to benchmark poverty lines is consistent with the definition of poverty used in this paper, since we are comparing the incomes of people at risk of poverty with a ‘middle’ income benchmark rather than low or high incomes. • By contrast, the average or mean income measure is more sensitive to changes in low and high incomes. Since these thresholds are on the face of it arbitrary, they need validation to assess whether they are acceptable to the public, whether they yield credible results both in aggregate and when measuring poverty rates among different groups (taking into account the results of complementary measures such as deprivation indicators), and whether they reliably capture the impacts of changes in policy and economic conditions: As indicated in Attachment 1, the 50% of median income measure is widely used by Australian poverty researchers, along with some government agencies (e.g. the Productivity Commission and Fair Work Commission): Poverty measurement: A poverty and inequality partnership position paper 32 • In 2024, this was approximately $560 per week for a single adult living alone. It is difficult to contest the view that an income below this is inadequate. This poverty line sits somewhat below the minimum household budgets calculated in budget standards research. • The 50% of median income measure is reasonably sensitive to significant changes in income support payment levels and economic conditions (such as increases or reductions in unemployment). • When housing costs are taken into account the profile of people living below this poverty line is reasonably consistent with that of people who are most likely to be deprived of essentials or experiencing financial stress.55 We use the 50% of median income benchmark as our primary poverty threshold, so that people with incomes below this level are estimated to be living in poverty. There is a strong case for supplementing this frugal measure with a higher poverty threshold (such as 60% of median income) for sensitivity testing purposes, as it is unlikely that people are not at risk of poverty as soon as they have a dollar of income above 50% of median income (or any other valid poverty line). Within this two-tier framework of poverty lines, people with incomes between the 50% and 60% of median income poverty lines may be described as ‘at risk of poverty’.56 One drawback of poverty thresholds linked to movements in median or average incomes is that they are highly sensitive to shifts in the business cycle such as booms and recessions. • According to the ‘relative’ approach to poverty measurement, a poverty line should rise or fall with shifts in overall community living standards. • However, ideally it should not change too abruptly with economic conditions – for example public expectations around minimum living standards are generally sustained (i.e. they do not fall sharply) during recessions.57 Henderson poverty lines In the original 1966 poverty survey conducted by Henderson and colleagues, the poverty line for a couple with two children was benchmarked to the ‘basic wage’ plus child endowment (family allowance) payments.58 This combination reflected a long-established Australian approach to setting minimum incomes: • The ‘basic wage’ was originally intended to meet a family’s essential living costs and was later supplemented by child endowment. A second poverty line was set at 120% of the above line: • People living below this higher line were described as ‘rather poor’ while those below the lower line were described as ‘very poor’. • A set of poverty lines for families of different sizes were derived from these two poverty lines, using an ‘equivalence scale’. Separate (higher) poverty lines were developed for households with a family 33 member participating in the labour force, to account for additional costs associated with employment or job search. A significant innovation was the calculation of a separate set of ‘after-housing costs’ poverty lines (as previously discussed), by deducting an estimate of typical housing costs from the above poverty thresholds (discussed in more detail later). These lower poverty lines were estimates of the income people needed to cover all costs apart from housing. To assess whether each household was in poverty, its housing costs were deducted from its income before comparing it with the after-housing costs poverty line. Despite the early connection with the Basic Wage, the link between the Henderson poverty line and actual minimum wages was tenuous from the outset and it was severed by the time the Poverty Commission reported in 1975: • During the 1960s the ‘basic wage’ fell out of use for wage fixation purposes and by 1966 actual minimum wages were significantly higher.59 • The Poverty Commission instead set what became the benchmark ‘Henderson Poverty Line’ (for a couple with two children) at 56.5% of average earnings.60 • In the 1980s when two-income families became more common, the Henderson Poverty Line was instead indexed to movements in household disposable income per capita. A key advantage of this approach to indexation is that the poverty line moves broadly in line with household living standards and can be updated each quarter using Australian National Accounts data.61 However, its original connection with Australian minimum income standards has faded over time. A further limitation of the Henderson Poverty Line is that it does not facilitate international comparisons. Budget standards As discussed previously, another way to benchmark incomes to a minimum acceptable living standard is to calculate the minimum household budgets required by different types of families to purchase essential goods and services:62 • Australian budget standards research uses a combination of normative judgements (e.g. the minimum cost of a healthy diet) and analysis of actual spending patterns of households with low incomes. • Focus groups of people assessed as at risk of poverty may be used to identify and validate the list of commodities included in each budget area to improve understanding of how people with low incomes budget. • The outcome is a set of minimum weekly expenditures required by different family types to meet their needs. Together with a margin for saving, these can be converted into disposable income thresholds to measure poverty. • A particular challenge in this type of research is the wide variation in housing costs across Australia. The key advantage of the budget standards approach is its transparency. A Poverty measurement: A poverty and inequality partnership position paper 34 measure grounded in the actual expenditure needs of families is likely to be widely accepted, especially since the budgets can be adjusted to reflect both expert judgements and public views and standards. • Budget standards can be used to validate income-based poverty measures such as those discussed above, including the equivalence scale used to compare the living standards of different sized households. • The main drawback is that, unlike the above relative poverty lines, regular, substantial research investment is required to update the budgets approximately every five years to ensure the budgets remain relevant to changes in community expectations and the cost of different items. A comparison of different poverty lines Figure 1 below compares different poverty lines used in Australia (in the case of a single adult living alone) with an estimated minimum budget for a single adult derived from budget standards research: • The 50% of median income (‘before housing’) poverty line sits between the Henderson (not in workforce) and Henderson (in workforce) poverty lines. • The 50% of median income (‘after-housing’) poverty line is lower than the equivalent ‘before housing’ poverty line since median housing costs are deducted. • The 60% of median income poverty line is higher than the above poverty lines. • The budget standard for a single adult is higher than all of the above poverty lines, though the difference would be less than indicated here if the poverty lines were indexed forward to 2024. 35 Figure 1: Poverty Lines for a single adult compared with a minimum budget from budget standards research ($pw in 2022-2024) Sources: Naidoo Y, Wong M, Smyth C & Davidson P (2024), Material deprivation in Australia: the essentials of life. ACOSS & UNSW Sydney; Melbourne Institute (2022), Poverty Lines; Naidoo Y, Bradbury B & Sawrikar P (2024), Budget Standards for Child Support Research, Final Report, Social Policy Research Centre UNSW Sydney. Note: Poverty lines are for 2022 while the budget standard is for 2024. ‘Before housing’ measures can be compared with household disposable incomes while ‘after-housing’ measures can be compared with household incomes minus housing costs. The Henderson ‘in the workforce’ poverty line applies to people who are employed or unemployed. The Budget Standard is a minimum household budget for a single adult, unemployed and paying rent equal to the average of the lowest quartile (25%) of Sydney, Melbourne and Brisbane rents. The income unit For the purpose of poverty measurement, individuals are usually grouped into households, and poverty thresholds are adjusted to take account of the needs of different sized households:63 • This assumes that incomes and living expenses (e.g. rent and food) are shared among adults in a household so that everyone has the same living standard. • That assumption is not universally valid – for example unrelated individuals sharing accommodation may share housing costs but not other expenses Poverty measurement: A poverty and inequality partnership position paper 36 - though it is likely to be closer to reality than an assumption that incomes and expenses are not shared at all. • In addition to ‘households’ the ABS also uses a category called the ‘income unit’, which includes immediate family members but excludes other adults in a household (e.g. adult children or single people sharing accommodation). This allows researchers to separate the incomes of those other adults from that of the rest of the household to determine whether they are living in poverty. However the ‘income unit’ is not generally used in poverty research as it is simpler to group people into households, and ‘household’ is more readily understood. • A key problem using household income to allocate individuals across the income distribution is that this does not take account of disparities in incomes within households. Adjusting for household size (equivalisation) To take proper account of economies of scale, household incomes are equivalised (adjusted according to household size). There is no single ‘correct’ equivalisation scale that can account for varying needs across diverse households. • The equivalence scale most often used in Australian poverty research is the ‘modified OECD equivalence scale’ developed by an advisory group of poverty researchers from wealthy nations, which assigns a value of one to the first adult, 0.5 to each additional adult and 0.3 to each child under 14 years (though in Australian research 15 years is the threshold most commonly used). This scale is used in the ABS Survey of Income and Housing .64 • The OECD also uses an equivalence scale based on the square root of household size, which is a more arbitrary approach (e.g. it assumes a child ‘costs’ the same as an adult) adopted for simplicity. • The Henderson Poverty Lines use an equivalence scale based on a New York survey of family living costs from the 1950s,65 which is unlikely to be relevant to Australian household budgets today. There is a case for using different equivalence scales for before and afterhousing poverty lines (as discussed previously). As mentioned earlier, a different set of equivalence scales would be appropriate to measure poverty in First Nations communities, which have different kinship systems to the rest of the community.66 Recommendations: Poverty thresholds 5. The household should be the unit for income-based poverty measurement and poverty lines should be adjusted for household size. • The base ‘equivalisation’ measure for this purpose is the ‘revised OECD equivalence scale’. 6. The primary poverty threshold should be 50% of median equivalent household disposable income indicating that people with lower incomes are living in poverty. This should be supplemented by a higher set of poverty lines based on 60% of median income indicating a ‘risk of poverty’. 37 7. Budget Standards research, updated at least every five years, should be used to validate income-based poverty thresholds. 8. As a complement to the automatic and immediate updating of poverty thresholds to movements in median incomes, alternative indexation methods should be developed and tested: • such as rolling average measures to reduce volatility in periods of rapid growth or decline in incomes across the community. 3.5 Data gaps and reliability Income-based poverty research relies on accurate, relevant and timely national data on the distribution of household incomes, supplemented by data on the distribution of wealth and consumption. The main sources currently available to researchers are: • The Survey of Income and Housing (SIH) and Household Expenditure Survey (HES) from the ABS; • The Melbourne Institute’s HILDA survey; and • Administrative data (e.g. tax and social security) which is gradually being adapted for research purposes through the PLIDA database. Timeliness A major problem for poverty researchers with these data sets is a lack of timeliness: • The latest HES was conducted in 2015-16; • The SIH is only conducted every two years and release of unit record data is often delayed, meaning there is a significant lag between data collection and release; • Data on wealth and material deprivation is collected every 4 years and not in every annual round of the HILDA survey. This means that unless researchers use microsimulation techniques, they may be reporting on rates of poverty and contributing factors that occurred years ago. Data gaps National income surveys often fail to include or identify some groups in the community who are at risk of poverty, such as: • People in First Nations communities; • People in institutional care; • People who are homeless or insecurely housed; and • People who are refugees or temporary migrants. Poverty measurement: A poverty and inequality partnership position paper 38 Recommendations: Data availability and reliability 9. The government should commit to providing the resources required by the ABS to undertake comprehensive and regular household income and wealth surveys as well as expenditure surveys (or at the least surveys measuring deprivation and financial hardship), at least every two years: • The income and wealth surveys should have samples of sufficient size and diversity to capture the main groups affected by poverty, include detailed data on wealth and financial hardship, and be comparable over time. 10. (a) To inform the development of multidimensional poverty measures and measurement of poverty dynamics, the government should commit to adequately funding the collection of longitudinal data, such as the Melbourne Institute’s HILDA survey. (b) If not collected in regular ABS surveys, deprivation data should be collected within HILDA at least every two years, at the person level. (c) Government should continue to facilitate the use of administrative data via PLIDA, including by improving access to administrative data on wealth, access to community services and the circumstances of First Nations communities. 39 Poverty measurement: A poverty and inequality partnership position paper 40 Income-based poverty research should yield timely estimates of the overall rate of poverty, its incidence among different groups including children, the depth of poverty among those impacted, comparisons with benchmark incomes (including social security payments and minimum wages), international comparisons, the persistence of poverty, its main causes and the impact of public policies. Headcount measures and poverty gaps The number and percentage of the population in poverty is of significant public and public policy interest, even though (and perhaps because) it takes a major change in economic conditions or public policy (such as the COVID lockdowns and related income protections) to shift the dial much on these aggregate measures.67 While it attracts less public attention, aggregate measures of poverty are incomplete without a measure of the average depth of poverty - the poverty gap - among people living below a poverty threshold: • One problem with average poverty gap measures (for people below the poverty line) is that they are impacted by measurement error in cases where survey respondents greatly underestimate their income. An income-based poverty measure should ideally facilitate timely comparisons of poverty rates in Australia and other countries. Population groups affected Timely evidence on the risk of poverty and profile of poverty among different groups in the population is needed to assess our fairness as a nation and for policy development. This may include, for example, breakdowns of poverty among: • people of different ages, especially children; • people of different genders; • family type; • people relying on different income sources, especially social security payments; • labour force status; • housing status; • geographic location; • First Nations peoples; • people with disability; • migrants, including more recent arrivals; and • people with diverse ethnic and linguistic backgrounds more broadly. 4. Outputs of poverty research 41 Attachment 2: Poverty measures used in other wealthy nations Country Poverty measures Income thresholds & adjustments OECD % of median income 50% of median equivalent household disposable income [Square root of household size or modified OECD equivalence scale] European Union % of median income Severe deprivation Living in a ‘jobless’ household 60% of median equivalent household disposable income [modified OECD equivalence scale] Lacking key essentials Canada Minimum budget Low income (% of median income) Market Basket Measure Sufficient income for a minimum budget for a family of four 50% of median equivalent household disposable income Spends 20% more of income on food, shelter and clothing than average United Kingdom* ‘Absolute’ low income ‘Relative’ low income Combined measure of material deprivation (lack of essentials) and low income 60% of the median equivalised household disposable income fixed at 1998 values 60% of the median equivalised household disposable income 70% of median income and falling below a material deprivation threshold [modified OECD equivalence scale] New Zealand* % of median income before-housing-costs; % of median income after-housing costs; Material deprivation Poverty persistence 50% of median equivalent household disposable income 50% of median equivalent household disposable income (after deducting housing costs) Lacking key essentials Duration of poverty Poverty measurement: A poverty and inequality partnership position paper 48 Country Poverty measures Income thresholds & adjustments Ireland % of median income Material deprivation Consistent poverty 60% of median equivalent household disposable income Lacking key essentials Low income and material deprivation combined Notes: * For child poverty only. A total of 43 (mainly low-income) countries use a Multidimensional Poverty Index while 34 European countries use Eurostat’s At Risk of Poverty and Social Exclusion (AROPE) measure which includes relative income poverty, and measures of employment participation and material deprivation.74 49 1 Budget standards are minimum budgets to achieve a socially acceptable living standard. Material deprivation measures the number of essential items that no one in Australia should go without but lack because they cannot afford them. A Multidimensional Poverty Index is a measure of wellbeing across a number of dimensions such as housing health and employment. All are explained later. 2 Davidson, P; Bradbury, B; and Wong, M (2022) Poverty in Australia 2022: A snapsho t Australian Council of Social Service (ACOSS) and UNSW Sydney; Davidson, P; Bradbury, B; and Wong, M (2023), Poverty in Australia 2023: Who is affected Poverty and Inequality Partnership Report no. 20. Australian Council of Social Service and UNSW Sydney 3 Economic Inclusion Advisory Committee (2024), Economic Inclusion Advisory Committee 2024 report to Government, Department of Social Services, Canberra; United Nations (2024), Sustainable Development Goals, End poverty in all its forms everywhere. Webpage, accessed December 2024; Australian Treasury (2023), Measuring what matters statement Australian Government, Canberra. 4 This definition is drawn from the Irish government (ed.) (2007), National action plan for social inclusion 2007-2016: building an inclusive society. Dublin: Stationery Office 5 Economic Inclusion Advisory Committee (2024) op cit; United Nations (2024), op cit; Australian Treasury (2023), op cit. 6 An example is the UN Sustainable Development Goal to halve poverty as measured nationally by 2030. 7 See for example: Naidoo Y, Valentine K & Adamson E (2022) Australian experiences of poverty: risk precarity and uncertainty during COVID-19 . Australian Council of Social Service (ACOSS) and UNSW Sydney; UNSW Sydney; Bowman D, Porter D & Mabare M (2021) M aking ends meet – fostering security and dignity in tough times . Brotherhood of St Laurence. Melbourne. 8 Expert Group on Poverty Statistics (2006), Compendium of best practices in poverty measurement. Rio de Janeiro; Alkire S & Santos M (2014) op cit. 9 Henderson R et al (1970), People in poverty - a Melbourne survey . Institute of Applied Economic and Social Research, University of Melbourne; Commission of Inquiry into Poverty (1975), First main report. Commonwealth of Australia, Melbourne 10 Saunders P (2015), ‘Closing the gap: the growing divide between poverty research and policy in Australia’. Australian Journal of Social Issues Vol. 50 No. 1, pp. 13-35. 11 This requires us, by 2030, to reduce at least by half the proportion of men, women and children of all ages living in poverty in all its dimensions according to national definitions, but no national measure has been adopted in Australia to track progress towards this goal. See also; Alkire S & Santos M (2014), op cit. 12 A total of 43 (mainly low-income) countries use a Multidimensional Poverty Index while 34 European countries use Eurostat’s At Risk of Poverty and Social Exclusion (AROPE) measure which includes relative income poverty, a measure of lack of employment, and/or severe material deprivation (Alkire S & Dirksen J (2024), Poverty in All its Dimensions according to National Definitions: A briefing on SDG Indicator 1.2.2, University of Oxford, OPHI Briefing 58/2024.) 13 Irish government (2007) op cit. Endnotes Poverty measurement: A poverty and inequality partnership position paper 50 14 Commission of Inquiry into Poverty (1975), op cit, p. 2. 15 Expert Group on Poverty Statistics (2006), op cit.; Alkire S & Santos M (2014), op cit; Hunter B & Biddle N (2017), Survey Analysis for Indigenous Policy in Australia: Social Science Perspectives . Ch.10. ANU ePress, Canberra. 16 Gordon D (2006), ‘The concept and measurement of poverty’, in Pantazis C, Gordon D & Levitas R (2006) Poverty and Social Exclusion in Britain , Bristol, The Policy Press. 17 These poverty thresholds are usually set at the household level, with individuals defined as living in poverty if they live in a household with economic resources less than the threshold (the poverty line) for that household type. In other approaches, such as the Multi-dimensional Poverty Index, the individual is the main unit of analysis. Alkire S, Kanagaratnam U & Suppa N (2021), T he Global Multidimensional Poverty Index, Methodological note 5 1, University of Oxford 18 Naidoo Y, Bradbury B & Sawrikar P (2025), U pdated Budget Standard Estimates for 2024, Final Report Social Policy Research Centre, UNSW Sydney 19 Naidoo, Y., Wong, M., Smyth, C. and Davidson, P (2024), M aterial deprivation in Australia: the essentials of life , Australian Council of Social Service (ACOSS) and UNSW Sydney; Saunders P, Naidoo Y & Griffiths M (2008), ‘Towards New Indicators of Disadvantage: Deprivation and Social Exclusion in Australia’. Australian Journal of Social Issues Vol. 43 No. 2, pp. 175-194 20 United Nations Department of Economic and Social Affairs (2024) op cit; Alkire S, Kanagaratnam U & Suppa N (2021), op cit. 21 Stewart K & Roberts N (2019), ‘Child Poverty Measurement in the UK: Assessing Support for the Downgrading of Income-Based Poverty Measures’. Social Indicators Research Vol. 142, pp. 523–542. 22 Alkire S, Kanagaratnam U & Suppa N (2021), op cit. 23 Financial stress is self-reported and refers to a lack of resources to meet current needs. There is some overlap between this and the broader deprivation research described above (for example, the inability raise a certain amount of money in an emergency is a variable that may be used in both). Phillips B & Narayanan V (2021), Financial Stress and Social Security Settings in Australia. ANU Centre for Social Research and Methods. Canberra. 24 This information can also be used to inform other types of poverty research, for example to set poverty thresholds or identify domains of living standards for inclusion in deprivation research. See for example: Naidoo Y, Valentine K & Adamson E (2022), op cit; Bowman D, Porter D & Mabare M (2021), op cit. 25 Townsend P (1979), Poverty in the United Kingdom . London, Allen Lane and Penguin Books. 26 Saunders P & Naidoo Y (2020), ‘The overlap between income poverty and material deprivation: sensitivity evidence for Australia’. Journal of Poverty and Social Justice, Vol. 28 No. 2, pp. 187–206. 27 Seebohm Rowntree B (2000), Poverty A Study of Town Life, Centennial Edition. Policy Press, Bristol. 28 Redmond G (2008), Children’s perspectives on economic adversity: A review of the literature. SPRC Research Discussion Paper No. 149 . Australian Research Alliance for Children and Youth (ARACY), 2014. The Nest action agenda: Technical document. ARACY, Canberra. Lee Y (2011), ‘The Effects of Persistent Poverty on Children’s Physical, Socio-emotional, and Learning Outcomes.’ Child indicators research , Vol. 4, pp. 725–747. 51 29 Hunter B & Biddle N (2017), op cit; Lahn J (2012), ‘Poverty, Work and Social Networks: The Role of Social Capital for Aboriginal People in Urban Australian Locale’ Urban Policy and Research , Vol. 30, No. 3, pp. 293–308. 30 Hunter B (2012), ‘Is Indigenous poverty different from other poverty?’ in Hunter B & Biddle N (Editors), Survey Analysis for Indigenous Policy in Australia. Centre for Aboriginal Economic Policy Research, Research Monograph No. 32, Australian National University, Canberra; Australian Institute of Family Studies (1993), ‘Aboriginal Australians and poverty, issues of measurement’ Family Matters, August 1993. 31 Closing the gap in partnership (2020), C losing the gap targets and outcomes 32 In 2009, the ABS set out to develop a conceptual framework of ‘low consumption possibilities’ based on income and wealth. The framework was intended to inform the development of measures to more accurately identify Australian households that have low consumption possibilities and are at risk of experiencing economic hardship. Consumption possibilities were defined as people’s command over resources that can be used to obtain goods and services to satisfy their needs and wants. (ABS (2009), Methodological news , June 2009). 33 Australian Bureau of Statistics (2022), S urvey of Income and Housing, User Guide : In this survey income includes receipts from: employee income (whether from an employer or own incorporated enterprise), including wages and salaries and other receipts from employment, income provided as part of salary sacrifice and/or salary package arrangements and non-cash benefits provided by employers, profit/loss from own unincorporated business (including partnerships), net investment income (interest, rent, dividends, royalties). government pensions and allowances, private transfers (e.g. superannuation, workers’ compensation, income from annuities, child support, and financial support received from family members. Household income excludes receipts from: capital transfers such as inheritance, lump-sum retirement benefits, life insurance claims (except annuities), compensation (except for foregone earnings), loan repayments, lottery and other gambling winnings, non-life insurance claims, receipts that result from a reduction in net worth (for example, sale of assets, withdrawals from savings, and loans obtained), holding gains/losses resulting from changes in the value of financial and non-financial assets and liabilities (for example, the value of shares held). 34 For example, from this standpoint savings (increases in wealth) should be counted as part of income but not consumption. So, employer contributions to superannuation should be counted as part of income (but is not in the ABS Survey of Income and Housing), and regular superannuation pensions should not be counted as income (but they are). 35 In the above definition of income, ‘current receipts’ generally refer to income over a period between a week and a year United Nations Economic Commission for Europe (2011), Canberra Group Handbook on Household Income Statistics. New York. 36 Oliver R (2018), Background paper – the concept of income . New Zealand Tax Working Group, Wellington. 37 Access to assets that can be drawn down is taken account in determining eligibility for some social security payments, through a ‘liquid assets test’, though this is controversial as it deprives people of savings buffers they are likely to need in the near future if fully reliant on income support. Poverty measurement: A poverty and inequality partnership position paper 52 38 Naidoo Y, Wong M, Smyth C & Davidson P (2024), op cit. 39 The annuitisation approach is also used in social security income tests through the calculation of deemed rates of return from assets. Relevant literature: Naidoo Y (2019), Comparing the implications of expanded income-based measures of living standards with an application to older Australians, Journal of Social Policy , Vol. 48, No. 1, pp. 83–105; Naidoo Y (2017), The standard of living and well-being of individuals – A case study of older Australians . UNSW Sydney; ABS (2021), Analytical measures of income, consumption, saving and wealth. Australian System of National Accounts: Concepts, Sources and Methods . ABS, Canberra. In a recent study of income inequality, the Productivity Commission used a measure it called ‘wealth-adjusted income’ which included the annuitised value of household wealth. As the Commission points out, this approach is more appropriate to measure income over a lifetime rather than a single year. (Productivity Commission (2024), A snapshot of inequality in Australia. Canberra). 40 The ABS warns that: there is an important debate in the international national accounts community over the validity of treating real holding gains in the same manner as gross disposable income. The impact of real holding gains on economic activity may not be equivalent to income received in cash or in kind. If a real holding gain accrues due to an increase in the price of a particular asset, many agents may wish to realise this gain. If they attempt to cash out at the same time, the price of the asset may fall and the size of the realised gain may be smaller than the imputed real holding gain. In addition, the realisation of a holding gain may lead to the payment of tax, which would reduce the amount of funds available to the asset holder. Thus, the measures introduced here should not be seen as replacing or correcting the traditional income and saving measures in the national accounts; rather they are provided to give users an alternative view of the available information. ABS (2021), op cit. 41 According to the Canberra Group of international experts on income measurement: one way to take account the impact on people’s spending power is to: ‘annuitise the net worth held by the household and add this (notional) annuity to the flow of income. However, annuitisation of net worth requires that a number of value judgements and assumptions be made in relation to, for example, the period over which the net worth should be annuitised (life of the householder or spouse) and the rates of return to be used. However, there are also simpler, but less sophisticated, methods available to use distributional information for income and wealth together.’ (United Nations Economic Commission for Europe (2011), ibid, p.4) 42 This would be an age beyond which people can reasonably be expected to start drawing down their superannuation (and superannuation rules allow them to do so). This could, for example be 65 years or the age of eligibility for Age Pension (67 years). 43 Naidoo Y (2019), op cit; Cigdem-Bayram M & Vera-Tosca E (2024), op cit. 44 Cigdem-Bayram M & Vera-Tosca E (2024), op cit. 45 Goods and services produced within households include imputed rents from owner-occupied housing (which can be captured indirectly through the ‘afterhousing’ poverty measures discussed later) and the value of unpaid domestic labour which is of wider interest from the perspective of gender analysis. 46 ABS (2028), Government Benefits, Taxes and Household Income, Australia methodology 53 47 Bradbury B & Saunders P (2021), ‘Housing costs and poverty: analysing recent Australian trends.’ Journal of Housing and the Built Environment , 15 September 2021. 48 Imputed rent measures do not take direct account of the capital value of the home. See ABS (2023), Survey of Income and Housing, User Guide, Australia 49 This is the main problem with poverty measures based on actual expenditure rather than income or spending power. One way to resolve this problem in regard to housing costs would be to cap deductions from household incomes to account for housing costs, especially mortgage interest payments. More people with large mortgage payments (who may well face financial hardship but are not living in poverty as commonly understood) would be excluded from the poverty count. 50 This is the main problem with poverty measures based on actual expenditure rather than income or spending power. One way to resolve this problem in regard to housing costs would be to cap deductions from household incomes to account for housing costs, especially mortgage interest payments. More people with large mortgage payments (who may well face financial hardship but are not living in poverty as commonly understood) would be excluded from the poverty count. 51 Mitra S et al (2017), ‘Extra costs of living with a disability: A review and agenda for research.’ Disability and Health Journal, Volume 10 No 4, pp. 475-484. Vu B (2020), ‘The costs of disability in Australia: a hybrid panel-data examination.’ Health Economics Review , Vol 10 No 6. 52 Ross R & Whiteford, P (1992), ‘Poverty in 1986: Aboriginal families with children,’ Australian Journal of Social Issues , Vol.27, No.2, May pp. 92-111. 53 Aguilar R et al (2024), March 2024 global poverty update from the World Bank. World Bank Group, Washington 54 OECD (2008), Growing unequal. Paris; Eurostat (2024), Glossary – At risk of poverty or social exclusion. Paris 55 Naidoo Y, Wong M, Smyth C & Davidson P (2024), op cit; Davidson P, Bradbury B, & Wong M (2022) op cit; Davidson P, Bradbury B, & Wong M (2023), op cit; Phillips B & Narayanan V (2021), op cit. 56 The European Commission describes the circumstances of people living below its 60% of median incomes in this way (Eurostat 2024, op cit). 57 The UK Social Metrics Commission have proposed an interesting method to deal with this problem, by benchmarking poverty lines to a rolling three year average value of median household incomes (Social Metrics Commission (2018), A new measure of poverty for the UK . London. See also Corak’s proposal for a five-year rolling average measure in Canada (Corak M (2016), Here’s a policyrelevant way to set the poverty line. ) 58 Henderson R et al (1970), op cit. The basic wage, introduced in 1907 in the ‘Harvester Judgement’ case, was the first minimum wage in Australia. It was indexed quarterly to the Commonwealth measure of inflation until 1953, when indexation was removed and replaced with increases through Basic Wage Inquiries. (Hamilton, R (2017), The History of the Australian Minimum Wage , Fair Work Commission). 59 Henderson’s original poverty line for a single adult was around half that for a ‘family of four’, hence well below the basic wage. This reflected the view at the time that minimum wages should meet the costs of a couple with children, and avoided any implication that (to eliminate poverty) income support payments for a single adult should be equal to the minimum wage. Poverty measurement: A poverty and inequality partnership position paper 54 60 This was based on the poverty line used in the 1966 survey, indexed forward to the 1970s. 61 Melbourne Institute (2024), Poverty Lines . Melbourne. 62 Naidoo Y, Bradbury B & Sawrikar P (2025), op cit; Saunders P, et al (1998), Development of indicative budget standards for Australia . Social Policy Research Centre UNSW; Bedford M, Bradbury B & Naidoo Y (2023), Budget Standards for Low-Paid Families . Report prepared for the Fair Work Commission, Melbourne, Australia. 63 The individual rather than the household remains the unit of analysis, but the household income of the person (adjusted for household needs) is used to measure their living standard. 64 ABS (2022), op cit, glossary 65 The 1954 ‘Family Budget Standard’ prepared by the Budget Stnadard Service of the Community Council of Greater New York was used to adjust the poverty line for different family size and composition. Stanton, D (1980), ‘The Henderson poverty line - A critique.’ Social Security Journal, December 1980, p. 18. 66 Hunter B & Biddle N (2012), ‘Survey Analysis for Indigenous Policy in Australia, Social Science Perspectives’. Research Monograph No. 32 , Centre for Aboriginal Economic Policy Research, Australian National University, Canberra. 67 Davidson P, (2022), A tale of two pandemics: COVID, inequality and poverty in 2020 and 2021. ACOSS/UNSW Sydney Poverty and Inequality Partnership, Build Back Fairer Series, Report No. 3; Sydney. Note that in most poverty research, this is expressed as the proportion of individuals rather than households in poverty, to better take account of the circumstances of people in larger households. 68 Fair Work Commission (2024), S tatistical report - Annual Wage Review 2023– 24 . Sydney 69 Harmer J (2009), Pension Review Report . Department of Families, Housing, Community Services and Indigenous Affairs, Canberra. 70 Economic Inclusion Advisory Committee (2024), 2024 Report to Government . Canberra 71 Treasury (2024), op cit. 72 Vera-Toscano E & Wilkins R (2022), The Dynamics of Income Poverty in Australia: Evidence from the HILDA Survey, 2001 to 2019 . Melbourne Institute. Cobb-Clark D (2019), Intergenerational Transmission of Disadvantage in Australia. Life Course Centre Working Paper No 2019-19, School of Economics, University of Sydney. 73 See for example: Naidoo Y, Valentine K & Adamson E (2022), op cit; UNSW Sydney; Bowman D, Porter D & Mabare M (2021), Making ends meet – fostering security and dignity in tough times . Brotherhood of St Laurence. Melbourne. 74 Alkire S & Dirksen J (2024), op cit. 55