Introduction

The costs associated with bringing up children are of intense interest to families and policy makers - the former because they bear most of these costs, and the latter because costs influence a range of family policies. However, producing robust evidence on this topic has proved challenging, in part because there is no agreed method for estimating these costs and the different methods that have been used are each subject to weaknesses.

The most common approach examines detailed data on how families spend their incomes. By comparing the spending levels and patterns of those with and without children, it is possible to deduce how the presence of a child (or children) influences the family budget and to ascribe this difference to the costs of a child (or children). This approach, while popular, must be able to unravel the difference between how much families choose to spend on their children and how much they need to spend. There is little sense in equating the former with the cost of children, since the notion of "cost" implies an element of necessity (associated with meeting the needs of the child). This cost is likely to differ from the choices that families make about how much to spend, since spending will be determined primarily by how much discretionary income is available, not by what the child needs. Children in "rich" families would therefore be identified as costing more than children in "poor" families. If policies were designed to reflect this difference, it would end up entrenching existing inequalities without appropriately addressing the underlying needs.

Alternatively, the budget standards approach starts with the needs of the child rather than the expenditure of the family. The latter is then derived by adding up the costs involved in meeting the needs of children and other family members and a family budget is derived that will allow the necessary items (and activities) to be purchased. This approach was first applied over a century ago by Seebohm Rowntree (1901) in his study of poverty in York, England, where budgets were derived to identify how much was needed to achieve "merely physical efficiency" - the basis of the poverty line used by Rowntree. In Australia, the approach was adopted by Justice Higgins, who used it to set the basic wage for a working family in the Harvester Decision of 1907.

After decades of neglect, the budget standards approach has been revived, initially following the important UK study by Bradshaw and others in the 1990s (Bradshaw, 1993) and reinforced more recently by contributions from a range of UK researchers including Morris, Deeming and Hirsch (Morris, Wilkinson, Dangour, Deeming, & Fletcher, 2007; Morris & Deeming, 2004; Deeming, 2005, 2010, 2011; Hirsch, 2015). The European Commission recently commissioned and released a major reference budgets report further highlighting the growing interest in the topic and its relevance to social policy design and development (Goedemé, Storms, & Van den Bosch, 2014; Goedemé, Storms, Stockman, Penne, & Van den Bosch, 2015). The term "reference budgets" is now used in the European context, although the Australian literature uses the term "budget standards", which will be used here to avoid confusion. The influence and application of the budget standards approach at a national level now extends to many countries, including France, Ireland, Italy, Japan, Korea, the Netherlands, Portugal and Spain.

These developments are underpinned by three related factors: first, the enduring appeal of the budget standards approach, which basically mirrors how actual families go about the process of budgeting to meet their needs; second, developments in research methodology that have addressed some of the weaknesses identified in earlier studies; and third, the availability of better data and analytical techniques that have allowed more robust estimates to be produced and made it easier to tailor the budgets to suit specific applications. Despite these important features, the approach remains subject to a number of caveats (discussed further below), which suggests that it can only provide a guide to action and, where possible, it should be accompanied by other evidence before specific decisions are made.

The main use to which budget standards have been put is in assessing the adequacy of incomes, normally minimum incomes such as those that form part of the social safety net: basic levels of pensions and allowances, family payments, the minimum wage and so on. This requires the term "adequacy" to be defined, and following the Harmer Pension Review an income can be identified as adequate when it provides "a basic acceptable standard of living, accounting for prevailing community standards" (Harmer, 2009, pp. xii-xiii). This raises complex questions about the meaning of "acceptable" and "prevailing community standards" that must be given clear articulation if relevant evidence is to be produced. This presents a formidable challenge to any research designed to develop an adequacy standard. Only the budget standards approach addresses these issues head-on.

Other approaches choose instead to either ignore the issue altogether or make a universal judgement - for example, that an income equal to one-half of the median is required to avoid poverty, or that a wage less than two-thirds of the median wage can be used to identify the working poor. Whether or not a specific income level or wage rate is adequate then depends solely on the implicit adequacy judgement without any supportive evidence. In contrast, the budget standards approach seeks to identify what is acceptable in terms of actual living standards and what prevailing community standards are and builds up the budgets from these understandings. This approach provides a far sounder basis for providing an independent assessment of income adequacy because it does not depend on an arbitrary judgement about the meaning of adequacy. Decisions still have to be made as the budgets are constructed and because many of these will be disputed, the budget standards approach is not a panacea. It does provide a flexible but focused method, however, for producing the evidence required to assess income adequacy independently and should therefore form part of the decision-maker's toolkit.

The remainder of this paper describes the main elements of a recent budget standards study conducted by researchers at the Social Policy Research Centre (SPRC) at the University of New South Wales. The focus is on what the new estimates imply for the costs of children although this is only one of a broad range of uses to which the new estimates can be put.¹ The next section provides a brief summary of the methods used to produce the new budgets, followed by a section presenting the new budgets and uses the estimates to derive the costs of children, compares them with earlier estimates using the same approach and discusses the implications of the findings. The main conclusions to date and areas for further work are summarised at the end.

Developing the new budget standards

A budget standard indicates how much a particular family living in a particular place at a particular time needs in order to achieve a particular standard of living. It is derived by specifying the standard that the budget is intended to support and then identifying and pricing every item that is needed by the family and each of its members to achieve that standard. Since budgets are normally derived for a range of family types, it is important to ensure that each is designed to achieve the same standard.

Developing the new budget standards involved only including items that were necessary to guarantee that all family members could achieve a full and healthy life, albeit one that involves a minimal level of outlays. The budgets for each individual were developed first, and those for extra family members (adults as well as children) were allocated similar items, while jointly consumed family items (e.g., the family home or items of furniture) were adjusted to reflect the change in family size or composition. If this equality of living standards is not achieved, then calculating the differences between the budgets for different families will conflate the cost of achieving a given standard with that involved in moving between standards. This point is important in the current context because the costs of children are normally estimated using the "difference method", which estimates the costs of children by taking the difference between the budgets for families with differing numbers of children, as explained further below.

Starting from scratch to construct a set of family budget standards is a daunting exercise. In order to minimise the effort (and cost) involved, the starting point for the research described here was the budget standard estimates produced by the SPRC in the 1990s (Saunders et al., 1998). The research that produced those budgets was commissioned by the federal government's (then) Department of Social Security and the project benefited from the advice and input of a range of experts in all aspects of family and household budgeting, nutrition, health economics and consumer behaviour.

Although the estimates are now over two decades old (the original budgets were priced in 1995), they are still regularly updated to reflect movements in the Consumer Price Index (CPI) by a number of community organisations, who continue to use them to inform adequacy judgements in a number of settings (see, e.g., Australian Council of Social Service [ACOSS], 2012, 2014). One area where the updated budgets have been used regularly is in the annual minimum wage setting process, where a number of submissions have used the estimates to argue for increases. This reflects the widely shared view, expressed in a recent submission to the minimum wage panel, that:

Indisputably, the best evidence in Australia about the needs of low income families is in the budget standards research of the SPRC. (Lawrence, 2015, p. 127)

Against this, the CPI adjustment is clearly a weakness because it effectively assumes that, aside from price rises, "prevailing community standards" are the same today as they were in the mid 1990s and this seems highly implausible. Reflecting this weakness, the Fair Work Commission (2014) has noted that:

We accept that contemporary budget standards measures can provide an effective means of measuring the needs of the low paid, which can be considered together with other relevant data. However, the budget standards measures derived from the 1997 (sic) SPRC study do not provide useful contemporary information about the needs of the low paid. (para. 390 [italics added])

The clear implication is that there is an urgent need to review the original budgets to determine what changes other than price uprating are needed for them to maintain their relevance. This was the motivation for the current study.

In theory, this was a straightforward (if tedious) process that involved examining the detailed item-by-item spreadsheets that underpinned the original budgets and repricing each item using current prices, or replacing the item if it was no longer available. In practice, what started out as a simple exercise soon became complex for several reasons: first, as the original budgets were reviewed it became apparent that there were many instances where the item itself needed to be modified or changed completely rather than just repriced. This was partly to reflect changes in "prevailing community standards" (particularly in the area of communications technology - the original budgets did not include a mobile phone for anyone!), but also to reflect changes in the options available to consumers (e.g., the rapid growth in "home brand" or generic items). There were also many instances where the assumptions built into the original budgets no longer seemed appropriate, particularly in relation to the lifetimes assumed for longer-lasting items, and how these vary with the circumstances of the family. Finally, a small number of errors were discovered, and these had to be corrected.

Due to the complexities mentioned above, the process of reviewing, revising and repricing the original budgets took far longer than originally anticipated. The original pricing of the food, clothing and household goods budgets began in the second half of 2013 (using as before, nationwide stores such as Woolworths and Kmart) but the new budgets were not finalised until over two years later, in early 2016. This process of review and revision was unexpectedly long because of the necessity to ensure that the revised budgets maintained their consistency both horizontally (i.e., between the different family types at each standard) and vertically (i.e., between the low-paid and unemployed standards themselves).

The process resulted in a set of budgets that had been reviewed far more carefully than was originally intended, so while the time taken was longer, the quality of the output produced was higher. The goal was to ensure that the new budgets embody the best of both international know-how and Australian practical experience and are therefore able to fulfil the high expectations of users as exemplified in the earlier quotations.

The process of review and revision described above identified numerous instances where the items themselves as well as their prices and lifetimes had to be modified. This painstaking process of reconstructing core components of the budgets was informed by three key elements. First, available data (mainly ABS survey data) was used to ensure that the budgets conformed broadly to what Australians actually have and do and what they spend their money on. Second, and importantly, a series of focus groups were held with the two groups that are the focus of the new budgets - low-paid working and unemployed individuals and families - to ascertain how they manage on their low budgets and, in particular, what economising strategies they employed to make ends meet.

The feedback provided by these focus groups proved to be extremely important and resulted in many changes to the budgets to reflect what was learnt. For example, focus group participants discussed the importance of "clothes swapping" and how this enabled them to get "new" school and other general clothing on a low income for children who had out-grown their own clothes. While an explicit allowance was not included in the budgets to allow for this, the lifetimes of certain clothing items were extended to reflect it. Another change that reflected comments made by the focus group participants was the inclusion of mainly "home brand" or generic brand items in the budgets. Participants suggested that these items were crucial because they were less expensive and enabled them to balance their budget.

Third, the accrued experience of the research team (two of whom were involved in the original SPRC study) with developing and using budget standards over the last two decades was valuable in helping to guide new decisions when they were needed.

One notable trend in the budget standards literature over recent decades has been the increased reliance on information and feedback provided by focus groups. This development reflects the greater constitutive role assigned to focus groups (particularly in the design of the UK minimum standards, where it has been argued that:

For society to agree a particular minimum standard of living, there needs to be informed negotiation and agreement about what constitutes a minimum, via a derivative of focus group methodology. (Bradshaw et al., 2008, p. 3 [italics added])

The counterpart to the increased role given to focus groups is a decline in the impact of "experts" who provide a range of information about family needs and what is required to satisfy them. This shift in emphasis is neatly summarised by Vranken (2010) as one in which the role of the focus groups has moved away from validating budgets designed by experts to one in which the focus groups design budgets that the experts then validate. However, this new perspective has not been universally adopted and the EU reference budgets study cited earlier did not follow the UK in assigning greater weight to focus group input, preferring instead to draw on information provided by an extensive network of national experts and informants.

In the current study, the focus groups were not asked to reach a consensus about the new budgets, only to provide feedback on preliminary estimates and suggest ways of improvement. Their role was also restricted by the practical challenges encountered in recruiting low-paid workers, who were difficult to get together in one place at one time due to the availability constraints of being in employment (often on a casual and/or irregular basis). Ensuring greater involvement of low-paid and casual workers in the construction of new budget standards is an important task for the future.

The standard applied to the new budgets is the minimum income for healthy living (MIHL), developed by Morris and other public health researchers in the UK (see Morris & Deeming, 2004; Morris, Donkin, Wonderling, Wilkinson, & Dowler, 2000; Morris et al., 2007). The approach involves a four-stage process: (1) draw on available public health research to identify personal needs in key areas of health for particular population groups; (2) translate this information into ways of living using existing surveys of lifestyles and public opinion; (3) cost these lifestyles in current circumstances; and (4) derive the out-of-pocket costs to individuals, after allowing for prevailing public provisions.

The approach has broad appeal because it specifies the targeted standard in a comprehensible way that few can disagree with: the concept of healthy living is now widely used to promote everything from dietary guidance to the need for appropriate regular exercise, the use of sunscreen, wearing appropriate clothing and footwear, and access to relevant facilities in the home, workplace and other institutional settings. The list illustrates the broad scope of the healthy living concept and highlights why it can be used as a targeted standard in the current context.

Although the budgets derived here cannot be claimed to be as firmly underpinned by public health research as is the case for the UK research cited above, it is important to note that all four words in the MIHL concept played an important role in guiding the research: Minimum - the focus is on identifying the minimal level of consumption that is consistent with the underlying ideal; Income - the focus is on how much money is needed each week to purchase the necessary items; Healthy - this is the key attribute that determines the items that are needed to ensure its attainment; and Living - which puts the focus on how people lead their lives and what is needed for them to attain and maintain the appropriate standard.

Throughout their development, the revised budgets were continuously reviewed to ensure their consistency with the concept of healthy living - as it affects people (adults and children) in their roles as consumers in the marketplace, as students at school, as parents in the home and as workers in the workplace. This focus on healthy living has the added advantage that it provides an important link between budget standards research and wider debates on social participation and inclusion, thereby increasing the relevance and value of the estimates in a range of public policy contexts.

It is also important to emphasise that the new budgets were constructed on a very conservative basis. In practice, this meant that when identifying the items to include and pricing them the approach that produced the lowest cost was always selected. This approach is in line with the focus on the minimalist nature of the MIHL concept, but also ensures that the budgets provide an appropriate benchmark for assessing the adequacy of the minimal incomes that form part of the social safety net. The overall budgets are therefore "tight" and there is no room to reduce them further without compromising the attainment of the MIHL standard.

The use of the MIHL standard represents a departure from the earlier SPRC study, which derived budgets at the Low Cost (LC) and Modest but Adequate (MBA) standards. These two concepts presented many challenges in the earlier study when trying to translate their definitions into practical budgets, and the decision to abandon them in part reflects these ambiguities. The MIHL standard provides a focus that extends across all areas of the budgets without being overly prescriptive about what the budgets should achieve, and this increased flexibility has made the task more manageable. In round terms, the new MIHL standard lies between the LC and MBA standards although exactly where it lies on that spectrum depends upon the cases being considered and the circumstances of the families to which the budgets apply. This will become evident as the discussion now shifts from methodology to findings.

The new budget standards and the costs of children

The new budget standards were derived for the following five family types: single people (females and males, grouped together into a gender-neutral budget); couple without children; couple with one child (girl, aged 6); couple with two children (girl aged 6 and boy aged 10); and a female sole parent with one child (girl, aged 6). Within each family, one adult (the male for convenience) was assumed to be either employed and being paid the minimum wage, or unemployed and receiving Newstart Allowance (NSA). These two situations reflect each family's dependence on one of the two main components of the Australian social safety net - the minimum wage and NSA. The research was designed to assess the adequacy of these two payment levels.

Because the focus of this paper is on the costs of children, the following discussion is restricted to couple families with zero, one and two children only. (The approach used to estimate the costs of children in sole-parent families is more complicated than that described below, for obvious reasons, and is not discussed further.) In each family, the prime breadwinner (for convenience the male parent with the exception of the sole family where it is the female parent) is assumed to be either working full-time on the minimum wage or receiving NSA, while his partner is either unemployed (if there are no children) or not in the labour force (where there are children). The partners of the unemployed men are also assumed to be unemployed and receiving the relevant social benefits.

Tables 1 and 2 present the new budget standards for low-paid and unemployed families respectively, with the costs of children estimated in each case by taking the difference between the relevant budgets. The "difference method" is most commonly used to estimate the costs of children within a budget standards context because it takes account of all the ways in which the presence of a child or children will affect the family budget. Some of these are directly related to the presence of the child, including the costs of the food they consume, the clothes they wear and the school expenses they incur. These costs are easiest to identify, and they formed the basis of the important study by Lovering (1984) that produced the first Australian estimates of the costs of children. However, they provide only a partial picture since they exclude all household items that are consumed jointly by all family members, adults as well as children. Only when all of these shared costs are included will the separate costs of each individual family member add up to the total family budget.

These other, shared costs are a legitimate component of the costs of children and include the additional energy bills required to keep the home adequately warm and the extra journeys that have to be made (and paid for) transporting children to where they need to be. Other shared cost items are even more obscure, including the extra costs imposed by the greater wear and tear that children impose on items of furniture and domestic appliances such as the refrigerator. In these instances, the assumed lifetimes of the item is lowered when there are children present and this raises the weekly cost (since the purchase price is averaged over a shorter period) relative to those without (or with fewer) children, so that the difference will indicate the extra costs of the child(ren).

The budgets shown in Tables 1 and 2 include an estimate of housing costs that is derived from the weekly rental data produced by the Real Estate Institute of Australia (REIA, 2016). Each family has been assigned a rental dwelling that reflects its size and composition, dwellings were then located within specific suburb types in each capital city and the average rents paid for these dwellings has been derived for the three largest cities: Sydney, Melbourne and Brisbane. The approach has been adopted so that others can vary the assumptions used and see what difference this makes to the budgets and calculations based on them. To maintain their relevance, the new budgets have been updated to the June Quarter of 2016 in line with movements in the relevant component of the CPI (further details are provided in the full report, Saunders & Bedford, 2017).

It is important to emphasise that the cost estimates shown in Tables 1 and 2 refer only to children with the assigned characteristics, that is, to a 6-year-old girl and to a 10-year-old boy. It would not be appropriate to assume that the costs apply to all girls and all boys, or even to all one- and two-child families since the items included in the family budgets will change along with the characteristics of the children.

It should be noted that housing costs for the 6-year-old girl are zero because both the couple with no children and the couple with one child are assumed to live in the same type of dwelling - a two-bedroom unit. If instead, the couple without children were assumed to live (like the single adult) in a one-bedroom unit, this would raise the implied cost of the first child by $73.40 a week for unemployed families and by $76.70 a week for low-paid families (see Saunders & Bedford, 2017, Table 5.11). The zero transport costs for the 10-year-old boy in the low-paid family reflects the assumption that his travel needs are fully accommodated within existing car trips and therefore involve no extra cost. In contrast, the unemployed couple with children has no car and must rely on public transport and therefore incurs extra transport costs in order to meet the needs of the older boy.² Education costs are zero for the couples with no children because the education costs relate only to children and the budgets include no allowance for adult education.

If the difference method is used to estimate the costs of children, the estimates in Tables 1 and 2 indicate that the weekly cost of the 6-year-old girl at the low-paid MIHL standard is around $137 a week (the difference between ($969.90 and $833.24), and that of the 10-year-old boy around $203 a week (the difference between $1,173.38 and $969.90). The combined cost of the two children (a better estimate in some ways since it averages out the costs that are specific to whichever of the two children is first) is (again using the difference method) $340 a week, or $170 a week per child. At the lower, unemployed MIHL standard these weekly costs are $106 (6-year-old girl), $174 (10-year-old boy) and $280 combined cost of $140 a week per child. In all cases the costs are lower, as expected, at the lower standard, with the difference amounting to about 18% when the costs are combined. The main contributors to these costs are food and housing, followed (in some instances) by education and transportation, although which budget areas contribute most to the overall costs varies across the budgets and family types.

Table 3 compares the new costs of children estimates with those produced in the earlier SPRC budget standards study and, for completeness, the partial cost estimates produced by Lovering (1984) over three decades ago. Attention focuses on how the new estimates compare with those based on the earlier study, updated in line with movements in the CPI between the March Quarter 1997 (when the original estimates apply) and the June Quarter 2016 (when the new estimates apply). The two sets of SPRC-based estimates vary in terms of the methods used to construct the detailed budgets and this compromises their comparability, although the aim of this exercise is to illustrate how relatively minor differences in method can produce estimates that differ greatly over an extended period when the impacts are allowed to cumulate.

What is clear from Table 3 is that the new estimates of the cost of children are considerably higher than those produced by updating the estimates produced earlier. This is despite the new budgets being deliberately kept to a minimum, as explained earlier. The reason for the large difference lies in the uprating process, which allows only for increases in consumer prices but makes no allowance for how "prevailing community standards" have changed over the period. These changes are captured in the new estimates (informed by the focus group feedback and other data used for benchmarking purposes) but are ignored in the updated estimates.

It is implausible to argue that community views on what represents the components of a minimally adequate living standard for Australian children have not shifted upwards over a period when general living standards have risen substantially. On this basis, the updated estimates have little relevance today. Instead, the new budget standards have been specifically designed to capture these changes and the estimates of the costs of children derived from them have similar legitimacy. Perhaps the single most important lesson to take away from Table 3 is the major biases that will be induced if budget standards are simply uprated in line with movements in the CPI, particularly when applied over extended periods.

Source: Saunders, 1999, Tables 1 & 2 and this article, Tables 1 & 2 above.

Main conclusions and future work

The budget standards approach identifies the key decisions, choices and assumptions involved in estimating how much is needed to achieve a specific standard of living. The MIHL standard that underpins the new budget standards is designed to be consistent with government policy goals in relation to meeting basic consumption needs, achieving healthy living and providing for an adequate level of social participation and inclusion.

The results provide an independent, evidence-based benchmark for assessing the adequacy of the incomes provided by key components of the social safety net including the minimum wage and Newstart Allowance, two of the key pillars of the income support system for working-age Australians. Here, they have been used to derive new estimates of the weekly costs of children, which are shown to vary between $137 and $203 for families in low-paid (minimum wage) work and between $106 and $174 for unemployed families in receipt of Newstart Allowance.

These cost estimates are far above those derived by uprating the estimates produced over two decades ago in line with movements in the CPI, and the difference illustrates the dangers involved in adopting an uprating method that does not allow for changes in prevailing community standards. Since this is a key feature of any reasonable definition of adequacy, its absence renders the estimates produced by CPI-uprating limited to the point of uselessness.

We do not claim that the new estimates alone should form the basis of a campaign to ensure that family payments and other components of the social safety net should be increased to cover the new estimates of the costs of children. However, the onus is on those who argue against such a move to demonstrate how families with children can meet prevailing community standards if they receive incomes that are below those implied by the new budget standards. Where can the savings be made without compromising the attainment of the MIHL standard?

A key advantage of the budget standards approach is this ability to examine the consequences of varying its underlying components and assumptions, one that no other approach shares. Of course, it is always possible to do better and this should be a longer-term goal. An increasing number of countries now see the merits of the budget standards approach and are funding research that will allow new estimates to be produced regularly, in the process expanding research capacity as a new generation of researchers acquires the skills needed to take on the task. As this process evolves, expertise will grow and areas of dispute will diminish as practice accumulates and "conventional wisdoms" emerge. For a country that relies more heavily than most on the income-testing of social benefits, it is difficult to fathom how Australia - once a world leader in budget standards research - has allowed itself to fall so far behind what others are now doing.

The new budget standards project represents a modest first step in addressing this anomaly and the results presented here illustrate how the research can guide current adequacy assessments and allow others to draw on the results in a variety of contexts where adequacy issues are central. These issues affect the living standards of all Australians and impact on the overall level of inequality. It is difficult to see how any level of economic inequality can be tolerated by society if basic needs are not met at an acceptable level, particularly for those at the lower end of the labour market or out of work. The underlying adequacy issues will not go away, and budget standards research can and should play an important role in helping to address them.

References

Australian Council of Social Service (ACOSS). (2012). Surviving Not Living: Submission to Senate Employment Committee on the Adequacy of Allowance Payments. Sydney: ACOSS.

ACOSS. (2014). Submission to Review of Australia's Welfare System. Sydney: ACOSS.

Bradshaw, J. (Ed.). (1993). Budget Standards for the United Kingdom. Aldershot: Avebury.

Bradshaw, J., Middleton, S., Davis, A., Oldfield, N., Smith, N., Cusworth, L., & Williams, J. (2008). A Minimum Income Standard for Britain: What People Think. York: The Joseph Rowntree Foundation.

Deeming, C. (2005). Minimum income standards: how might budget standards be set for the UK? Journal of Social Policy, 34(4), 1-18.

Deeming, C. (2010). The historical development of family budget standards in Britain, from the 17th century to the present, Social Policy and Administration, 44(7), 765-788.

Deeming, C. (2011). Determining minimum standards of living and household budgets Methodological issues, Journal of Sociology, 47(1), 17-34.

Fair Work Commission. (2014). Annual Wage Review 2013-14. Melbourne: Fair Work Commission.

Goedemé, T., Storms, B., & Van den Bosch, K. (2014). Pilot Project for the Development of a Common Methodology on Reference Budgets in Europe: Proposal. Brussels: European Commission Directorate-General for Employment, Social Affairs and Inclusion and University of Antwerp.

Goedemé, T., Storms, B., Stockman, S., Penne, T., & Van den Bosch, K. (2015). Towards cross-country comparable reference budgets in Europe: First results of a concerted effort. European Journal of Social Security, 17(1), 3-31.

Harmer, J. (2009). Pension Review Report. Canberra: FaHCSIA.

Hirsch, D. (2015). Inflation and the minimum income standard - past and future measures. Loughborough: Centre for Research on Social Policy, University of Loughborough.

Lawrence, B. (Ed.). (2015). Working Australia, 2015: Wages, families and poverty. Melbourne: Australian Catholic Council for Employment Relations.

Lovering, K. (1984). Cost of Children in Australia (Working Paper No. 8). Melbourne: Institute of Family Studies.

Morris, J. N., & Deeming, C. (2004). Minimum incomes for healthy living (MIHL): Next thrust in UK social policy? Policy & Politics, 32(4), 441-454.

Morris, J. N., Donkin, A. J. M., Wonderling, D., Wilkinson, P., & Dowler, E. A. (2000). A minimum income for healthy living. Journal of Epidemiology and Community Health, 54(12), 885-889.

Morris, J. N., Wilkinson, P., Dangour, A. D., Deeming, C., & Fletcher, A. (2007). Defining a minimum income for healthy living (MIHL): Older age, England. International Journal of Epidemiology, 36(6), 1300-1307.

Real Estate Institute of Australia (REIA). (2016). Real estate market facts: March quarter 2016. Canberra: REIA.

Rowntree, B. S. (1901). Poverty: A study of town life. London: Macmillan.

Saunders, P. (1999). Budget standards and the costs of children. Family Matters, 53(Winter), 62-70.

Saunders, P., & Bedford, M. (2017). New minimum income for healthy living budget standards for low-paid and unemployed Australians. Sydney: Social Policy Research Centre, University of New South Wales.

Saunders, P., Chalmers, J., McHugh, M., Murray, C., Bittman, M., & Bradbury, B. (1998). Development of indicative budget standards for Australia (Policy Research Report No. 74). Canberra: Department of Social Security.

Vranken, J. (2010). Using reference budgets for drawing up the requirements of a minimum income scheme and assessing adequacy: Synthesis report. Brussels: European Commission Directorate for Employment, Social Affairs and Inclusion.