Labour is there (USA) so well rewarded that a numerous family of children, instead of being a burthen is a source of opulence and prosperity to the parents…. A young widow with four or five children, who, among the middling or inferior rank of people in Europe…is there (USA) frequently courted as a sort of fortune (Adam Smtih 1776, 2003): 100)

1 Introduction

In this article we will argue that measures constructed on the foundation of empirical knowledge of consumption needs and consumer expenditures are the best approach to answer a very frequent question in public policy—“How much does it cost a family to live?”, or How much does it cost to raise a child?” (Fisher 2001).These questions may be seen as a practical expression of the recurrent question in welfare and poverty research: how to determine a socially acceptable standard of living? Further we will argue that traditional income approaches are less suitable to determine such standard of living because they end up with levels that are more or less incidental points on the society’s income distribution, without knowing the actual welfare consequences of living on these. For the children an additional consequence is that their needs are defined as a fixed relation to their parents, or more exactly to the “first” adultFootnote 1. The principal difference between these two approaches is—in other words—that consumption constitutes the material content of a given level of living while the income approach defines rather abstract income levels.

‘Relative’ and ‘absolute’ are often used to distinguish between consumption and income based measures. This distinction, however, does not encompass the principal differences between them. Both approaches we discuss in this article are relative in the sense that the level changes with the overall economic and social changes in a given society. This in contrast to an absolute poverty line that “is one which is constructed as an estimate of families’ minimum consumption needs; this is done without reference to the income or consumption levels of the general population” (Fisher 1995: 2). It is—in our view—more convenient to use the expression ‘substantial’ to measures that are constructed on the basis of consumption needs and consumption expenditures i.e. that have a specific material grounding. This does not mean that a ‘substantial’ measure can never be absolute. The poverty measure in USA is, for example, based on consumption, but does not “reflect changes in consumption patterns and changing concepts of what constitutes a minimally adequate standard of living” (Ruggles 1990 quoted from Fisher 1992: 1). The US definition is in other words absoluteFootnote 2 even if it has a material grounding. In this article, however, our primary concern is to discuss the applicability of the two relative measures—one substantial and one formal—in assessing the cost of children in a practical and political context.

We will further argue for a principal division between indicators used as monitoring devices on the one hand and indicators designed to inform policy on the other. Income measures of poverty and the utility-based equivalence scales are good examples of the former, while substantial poverty measures and standard budgets are examples of the latter. The distinction is often confused in the literature of poverty measures and equivalence scales, and this fact creates unproductive discussions about which measures are better suited without contextualising it. The cardinal argument against substantial poverty and standard budgets is that they are normative and thus unscientificFootnote 3. The alternative is either to reject the idea of defining levels of living altogether, or to hide the normative implications in pure statistical measures. Substantial measures—on the other hand—make explicit normative arguments. Efficient use of public funding and obligations towards people in need imply normative reasoning in combination with empirical knowledge of the relation between material wellbeing, cost structure and income. The purpose of this article is to contribute in the development of a measure that combines these elements.

Our principal argument is that standard budgets may be used as a substantial measure that more precisely reflects the economic realities of consumption units across different household types and social divisions than the traditional income approach. Hence we argue that an approach based on needs rather than income is a much better foundation for practical political initiatives to enhance social inclusion and equality (Borgeraas and Dahl 2007). There are several advantages to use the standard budget method to assess levels of living. First, it offers a measure that is very specific in terms of actual needs (Saunders 1999: 44)Footnote 4. Second, it can be used to define multiple levels of living—minimum, poverty, modest but adequate, and reasonable (Bradshaw 2001a, b)Footnote 5. Third, it may reflect nationally defined welfare ambitions, which in turn lend themselves to cross-national comparisonsFootnote 6. Fourth, standard budgets take children’s consumption needs as a staring point and not only as fractions of the household’s, or the adult’s or the nation’s income distribution.

2 Needs vs. Preferences or Policy vs. Description

The material welfare perspective is rooted in classical economic theory that defines economy as the science of material welfare (Cooter and Rappoport 1984). We may characterise the classical approach as materialistic in the sense that welfare is defined in terms of the material resources individuals or households have at one’s disposal (Collette 2000). In this tradition material possession is evaluated against human needs in a given society. A given quantity and quality of material goods to live a decent life and the income necessary for acquiring these goods, define the level of living. First it is, according to this theory, necessary to define needs for consumer goods, and then it is possible give a meaningful definition of the income level that is necessary for this level of living.

Modern welfare theory, on the other hand, is more concerned with income as a generalised expression for purchasing power. What individuals buy for their income is an expression of preferences and is seen as a given fact, and there is no scientific standard that can evaluate preferences. One serious consequence of this position is that income as welfare indicator is drained of content and becomes exclusively an expression of income differences. There is no doubt that, for example, 100 dollars have a higher purchasing power than 50 dollars, but this idea has to be qualified. The expression often used in economic models is ‘all other things equal’ but this is intellectual laziness and says nothing more than the purchasing power statement. It is not a linear relation between income and welfare and it seems obvious that 50 dollars have a greater impact on the welfare for people with low or no income than the welfare effect of 100 dollars for rich people. The basic idea guiding the discussion of defining poverty lines is that a given amount of money is more crucial for poor people than for the non poor, and that this sum of money gives a qualitative better welfare for those in poverty than for the rest. The importance of the households’ income for children’s welfare compared to other characteristics of the parents and the household is still unsettled (Mayer 1997).

In macro analysis income distribution is used as a measure for inequality. Simple statistical measures as deciles, quartiles, mean, median etc are used to express a picture of income dispersion. The interpretation of empirical income differences and the policy implications of these differences are, however, not a question of statistics. The absence of a material income concept makes it almost impossible to define levels of living that are not highly arbitrary and normatively unacceptable as a standard for policy. Income poverty is an excellent example.

2.1 Poverty

Poverty is genuinely a political concept (Simmel 1971). It is almost meaningless to define poverty unless one is prepared to do something with it. Many discussions of how to define poverty are—probably—hiding a political disagreement about who are worthy receivers. And it is reasonable to believe that the more specifically poverty is defined, the greater the possibility for political disagreement. Or—to turn the argument the other way: the more abstract the definition, the less the possibility for political disagreement will be. This may be the reason for the lack of political resistance against the income definition of poverty.

In the European Union poverty is defined as household income below 60% of the median income (Atkinson et al. 2002). This is an inequality measure, and is as such a convenient method for identifying levels of inequality at a given time, for monitoring the developments over longer time-spans and for comparing the level of inequality between countriesFootnote 7. The approach is, however, more problematic when it comes to policy action to solve poverty, for example money transfers to poor families. The crucial question is why people with an income slightly above 50% of the median income are not eligible for support? Why not define 45% of the median as poverty? Or—more principally; what criteria do we use to decide who are poor and who are not? The median is a statistical measure with no specific content, and can be used to characterise all types of quantities. The choice of a specific income poverty level must—in our opinion—be based on an empirical understanding of the standard of living that a given income can offer. If it is possible to have a decent life for most people with an income equal to 60% of median income, it will be inefficient—and unjustifiable—to design an anti-poverty policy for these groups. Or the other way around—if a substantial portion of individuals with 65% of the median income does not have a decent life, the 60% measure would also be unjustifiable. The point is; without a substantial concept of income we do not know the standard of living that a—more or less arbitrary—income level can offer. A minimum condition for a legitimate and efficient anti-poverty policy is that the poor are deprived of some resources that are necessary to satisfy some basic needs.

The arbitrariness of the EU poverty line can be exemplified by the consequences of changing the poverty definition from 50% of the median to 60%. Norway (not an EU member) decided—roughly speaking—to adopt the EU definition of poverty, which was at the time 50% of the median incomeFootnote 8. Using a 60% threshold in Norway would imply that pensioners with the minimum pension—a governmental pension to old people with no or too little income from the labour market—would be defined as poor. This is both politically unacceptable and probably offensive to the public opinion. There is also disagreement among social scientists and other experts whether or not people on minimum old age pension should be characterised as poor. If the Norwegian government accepted the new definition of income poverty, the consequence would either be to raise the minimum pension or officially recognise that the government’s support for this group keep them trapped in poverty, both of which is hard to do. To be useful as a practical measure it is of great importance that poverty is reserved for those in need. This implies that the levels of living of those who are targeted are known. A strict income measure has not this character.

2.2 Equivalence Scales

The same arguments can be raised against most of the so-called equivalence scales. The equivalence scales used for policies in OECD countries calculate the cost of the household members as shares of the first adult. Due to economy of scale each additional adult costs 50% of the first adult, whereas children under the age of 14 have a consumption need that is 30% of the first adult. This is the so called modified OECD scale. An earlier version of this scale was 0.8 for additional adults and 0.5 for children. The change of scale seems counter-intuitive given the fact that this happens in a period that children’s wellbeing, consumption (expenditure) and hence the cost of children have increased considerably (Miles 1998; Schor 2004; McNeal 1999).

The question we need to ask is; how much additional income do households with children need in order to be on the same welfare level as households without children?Footnote 9 A reasonable answer to this question is in terms of the consumption expenditure of children. Even if this is recognised theoreticallyFootnote 10, most of the literature of equivalence scales focuses on the reduction of the consumption of the adults before and after they get children. The income ratio between children and ‘additional’ adults are treated as a kind of zero-sum game in the sense that the cost of children reduces the adults’ welfare unless the cost is compensated for by a rise in income.

There are three serious objections to the traditional equivalence scales: The first objection is that this approach is not able to include life phases from a welfare level perspective. This is primarily due to the conventional economic assumption that individual welfare optimum is a situation where no-one can improve their welfare without lowering the welfare of others. In most cases childbirth boosts the household’s expenses without compensating rises in income. Given a specific income an additional child will therefore automatically reduce the other household members’ consumption, and hence the adults’ welfare. Much effort has been directed towards developing methods suitable for defining how much additional income the household needs in order to compensate for the rise in expenditures. We will argue that the theoretical assumptions in classical economic welfare analysis are not suitable for analysing children’s welfare, first of all because households with and without children are in two different life phases. It is simply unreasonable that the adult’s consumption levels and composition should be the same irrespective of whether he or she is supporting children or not. It is not reasonable to assert that parents continue to live and consume exactly in the same manner as they did before they had children. Child rearing in a modern society demands that parents change their social lives—less beer drinking, restaurant visits, holidays etc.

An alternative hypothesis can be that changes in parents’ consumption pattern as a consequence of childbirth often reduce the consumption expenditure equal to the additional cost of having the first child. This hypothesis implies that the economic burden for the household of having the first child is much lower than the direct cost of the child. Furthermore, in spite of the economy of scale of having more children, the costs of additional children are higher for the household than the direct cost. More generally, we argue that most of the models developed to compare the welfare between households with and without children lack a genuine child perspective. It is the adult’s welfare that is compared, not the children’s.

Children are not a homogenous social group. Being a ‘child’ will change through childhood when children pass through different phases that bring new challenges and activities. As they grow older they will spend more time away from home, and peers and school will get a more prominent place in their life. Consumption will increase because they need consumer goods to participate in different activities as well as the symbolic aspects ensuring that they fit in with their peers (Middleton et al. 1997, Brusdal 2004).

The second objection is that the scales are fixed. This implies that the consumption ratio between adults and children is the same irrespective of the households’ income, class position, ethnicity, nationality, historical period etc. What is more, if the households’ income is below the poverty line it seems meaningless to develop cost ratios between the members of the household. This is, however, a normative statement and is difficult to analyse empirically. On the other hand there are empirical studies that indicate that the scales differ substantially between household with different income levels. For instance Aaberge and Melby (1998) compared a household with two adults and two children with single households with different income levels. Using income dependent scales they found that—in order to be at the same welfare level—a four person household must have 191% higher income than a single person household with a yearly income of NOKFootnote 11 50,000. The ratio is 10% for a single household with NOK one million a year. Using a standardized scales system may work in macro analysis, but seems more dubious as a method for comparing poor with non-poor households, or as a guideline for identification of poor households.

Yet another point is that these scales do not only differ between households with different income levels. It seems reasonable to assume that there are cultural and ethnic differences in how parents prioritise children. Moreover, there are hardly any good arguments in support of the proposition that there are no significant national differences as to how much children are prioritised.

The third principal objection is that the underlying philosophy supporting the equivalence scales presupposes an almost extreme relative income position—income and preferences are given—and an extreme absolute position when it comes to social differences—differences between income groups are as they are. We go the other way around and advocate a view that takes social relativism more seriously and downgrades the income relativism. There are some income differences that are more devastating than others and there are income levels that are too low, regardless of the preferences, especially when it comes to adults’ translation of children’s preferences. This means that we argue for a position where the notion of income level must be “filled” with content—i.e. consumption items and services needed to support children’s basic needs and development. It seems that the experience with the so called ‘ordinalist revolution’—the rejection of cardinal utility and the possibility to compare utility across individualsFootnote 12—indicate that it is not suitable when it comes to developing policy measures that monitor children’s welfare. Children—as every individual—have needs, some more basic than others. Countries with welfare ambitions have to define these needs and estimate how much they will cost. In this perspective the income is the dependent and needs are the independent variable, and not the other way around as the ordinalist tradition seems to imply.

The arguments do not necessarily imply that income poverty is a useless measure. The point is more that income poverty may be suitable to monitor the development of poverty—or inequality—over time and to conduct cross-country comparisons. Another advantage of income poverty measure is the scientific legitimacy. Its political legitimacy is—however—probably lower, especially among authorities with direct responsibilities for the poor. In spite of the statistically precision, income poverty is almost useless in real life, first of all because it has little known content. Even if the conventional ordinalist economic theory rejects normative reasoning, income poverty is—as long as the word poverty is used—a normative concept even if the normative implications are hidden behind a purely statistical measure. The scientific objection is that in poverty research income must be treated as a dependent variable and not as an independent, as is the case for most income poverty concepts. If needs, level of living, and conceptions of acceptable/unacceptable living conditions come first, it becomes convenient to ask how much income is necessary to avoid the unacceptable living conditions. We are, in other words, asking for a poverty measure more in line with the classical economic tradition that started out with the material conditions and then asked how much income is necessary to be able to have a given level of material satisfaction.

3 Indicators to Inform Policy

In line with the classical tradition (see Cooter and Rappoport 1984) we have to ask some important questions before we can define a poverty line;

  • What do families and children need in different life phases?

  • What types of items (consumer goods) and activities can satisfy these needs?

  • How much does it cost to buy these items and to participate in these activities?

  • What income levels equal this consumption level?

We will discuss and launch the standard budget approach as an absolute measure that will estimate how much money each member of a household must have to sustain a certain standard of living. This approach ensures that every member of the household is treated as a unique individual and not as a percentage of a household budget.

3.1 Measuring Children’s Needs

Our focus is children’s needs in relation to their everyday life in order to sustain a certain standard of living. These are many and most of them are costs like expenses for food, school and learning, health, clothing, leisure activities etc. But children need other things as well. It is possible to divide children’s needs into family-related needs and child-related needs. Some of the family-related needs have a direct impact on children’s everyday life both today and in the future, like the family's living conditions. But modern childhood creates some specific needs for consumer goods that are only related to the child. In the following we will elaborate on this, but first of all, we will take a closer look at what it means to be a child in a commercialized world.

To be a child is not a role. It is influenced by biology, but not determined by it. For how long you are a child and how you live like a child is determined by economic and cultural patterns and differs between societies as well as within societies. Being a child means that you are on your way into adult society, but you are also present here and now. A child can be understood as a “being” with reference to the child’s situation and conditions here and now and their rights as a child. A child is vulnerable and dependent and subordinated to the family’s standard of living. The term “becoming” stresses that children are developing and their conditions for development influence their future. Both these aspects have to be considered when taking account of children’s standard of living.

Childhood changes through time. One characteristic of modern-day childhood is that it has become commercialized. Children have become consumers at a much earlier age than before and to a much larger extent. While emphasis has been put on the symbolic aspect of consumption, less attention has been paid to the functional one. However, consumer goods are often necessary in order to participate in activities and it is important for the individual’s access and social life. Nowadays, children play less in streets and parks, and most children participate in an organised activity. This costs money, both the fees and the equipment. Activities have both dynamic and symbolic aspects. The activity aspect offers participation and social relation with your peers, and the possibility to obtain competence from participation in the activity as well as social competence. Consumer goods are an important part of children’s activities at home as well, whether they are electronics or other toys. Miles (1998) writes that consumer goods have a fundamental effect on children’s everyday social life meaning that the family’s economy will influence the child’s life both here and now as well as in the future. Children in families with few economic resources will experience exclusion more often both today as well as in the future.

Another important aspect of consumption is the symbolic aspect. Society develops in the direction of greater individualisation and the staging of oneself is done with a consumer practice where the individual defines him-/herself through the signals sent out by material carriers. Material possessions have a symbolic significance for their owners, as well as for other people. Possessions play an important role in everyday life: signifying both who you want to be and to whom you belong, but also what you do not want to be associated with. This is of particular importance to children and youth trying to answer the question “Who am I?” According to Douglas and Isherwood (1979) consumption is necessary to make visible and stable categories of culture. To own and to do what your peers own and do is important and makes it possible to maintain a common basis of beliefs.

Consumer goods can be used strategically in an inclusion/exclusion process and as means to get a higher position in a status hierarchy. Molnar and Lamont (2008) show how low-status groups, black Americans, use consumption to express and transform their collective identity and acquire social membership in society or in subcultures. Consumer goods can thus be used in order to fit in and be part of a larger group, and conversely, a lack of consumer goods can cause bullying or exclusion (Ridge 2002). Having the “right” consumer goods might also be important for your self image.

Consumer goods are embedded in children’s everyday life, and their ability to consume influences their activities as well as how they see themselves and are seen by others. That means that their ability to consume is essential for their wellbeing.

3.2 Standards of Living-inclusion and Exclusion

Most of our basic needs are taken care of by consumer goods; we buy food when we are hungry and put on clothes when we freeze. However, consumption is culturally defined and it is more than just satisfying basic needs. The many aspects of consumption suggest that whether you have these things or participate in these activities or not might have some effect on your standard of living. Roseborough (1960) writes that facilities (consumer goods) are means in the performance act and necessary to be able to participate in different activities. Further, he writes that consumption has a latent function in that the possession symbolizes the fact that the possessor has the right to perform such acts and has an obligation to do so. The cost of these consumer goods will vary between different social systems, and so will their importance. However, some items are central in all systems. A person needs to possess a minimum of goods and participate in a minimum of activities, and by doing so this signals that he or she is an acceptable member of society. This is the standard package:

Assuming a cultural tradition, that there will be a complex of goods and services, a certain minimum of facilities, which each household will possess or at least will have use of, and of which will signify to others that the members of that household are acceptable members of society. This complex, in American society, has been called a standard package (Roseborough 1960: 463).

The standard package suggests an inclusion and exclusion process because if you do not have this minimum you are not able to participate in common activities. Being integrated in a peer culture means to be able to participate as equals that is to consume the same things as they do. Many studies have shown that parents’ economy influence children’s lives (Ridge 2002; Eydal 2006; Rysst 2006; Middleton et al. 1997) and children from poor families are more likely to experience exclusion. This could be an unpleasant experience, and an Icelandic study found that children from less affluent families, according to their own definition, expressed a greater longing for “things” than children from more affluent families (Eydal 2006). The wish to be integrated seems to be strong. The fact that most activities have a price tag and not all children have the money to “buy” will also result in fewer possibilities for acquiring different types of competence for some children (Rysst 2006; Chin 2001).

The symbolic aspects of consumption have consequences as well. Ridge (2002) focused on the need for “fitting in” and “joining in” when looking at children’s ability to participate in the social world with their peers and found three key arenas where economy affected their relational lives:

  • The capacity to fit in with their peers.

  • The importance of clothes as a key signifier of peer inclusion

  • The opportunities to join in shared social activities with their peers.

Children’s standard of living operates in a time span; children are both being and becoming. Their consumption is important here and now, but consumer goods can have a strong impact on their future life because of the competence the possession of special consumer goods might provide. Of current interest are activities and educational consumption that will enable the child to succeed later in life. The potentials for development are a central part of childhood and something we implicitly talk about when we defend the need for education, participation etc. What children are doing when they are young is considered in relation to the effect this has on their future life. What is called pedagogical toys and pedagogical activities are just characterized by their expected future effect.

3.3 From Standard Packages to Standard Budgets

We will argue that in almost every national culture there exists a widely held—but often very vague—concept of what it means to be a family. This concept is expressed and defined in terms of material goods and social activities. In line with Parsons and Smelser ([1956] (2001) we hold that the institutionalization of the family system…implies a certain minimum of possessions in order to meet the cultural (as opposed to mere legal definition) of a family (p. 222). However, in modern societies we can observe greater individualization and de-institutionalization of the national family system due to diffusion of multiculturalism based on ethnicity, age, gender, life styles etc. This de-institutionalization process contributes to the vagueness of the standard package concept and trigger active resistance to living like an ordinary or typical family. Nevertheless most people will have an idea of what a ‘typical’ family should possess in the national culture in which they live. This notion is also vital as a guideline in the public discussions of acceptable and unacceptable living conditions.

Even if it is possible to develop theoretical arguments for the existence of a popular concept of a typical national family and a subsequent consumption pattern, it is a huge step to define the specific content of this package and to set a price tag on it. In line with the classical poverty studies, one strategy is to develop a so called standard budget. This is a budget that expresses consumption levels in terms of consumer expenditures for households of different sizes and composition. The principal advantage of this strategy compared to the income approach is that the cost of having a consumption level equal to the standard package may be translated to the income level that is necessary to be able to buy these possessions. In contrast to the income approach the standard budget method defines the income level as a function of the consumption level and not the other way around.

Most commonly, standard budgets utilize an expert approach. This implies that experts in the different consumptions sectors are responsible for defining a given consumption level in terms of specific goods and services. This does not—of course—mean that it is up to the experts own understanding to choose the items. It has to be developed a set of common criteria—based on empirical knowledge of the consumption level and the distribution of specific items in various household types in the population. The existence of public norms concerning a specific consumption, for example national nutritional recommendations and knowledge of the population’s attitudes toward consumption has to be considered. The relation between empirical knowledge and more or less explicit normative regulations and attitudes and the end result is complicated. A realistic basket must be balanced between ideal demands and empirical reality. The experts’ task is to do this as explicitly as possible and thus be open for criticism. The basket is developed in this ‘communicative’ framing as should be the ideal of all measures that have implication for policy.

The cost of buying these items is the “budget” expressed in money. The Norwegian standard budget—as we will use as an example—consists of, in principle, a complete list of goods and services that are considered necessary to maintain a reasonable consumption level within a great number of consumption sectors. The actual costs of the defined items within each of these sectors add up to the consumer expenditure level.

In the Norwegian budget there are 12 consumption sectors that are divided into two main segments of expenses; individual—and household-specific. Individual-specific expenses are related to individual consumption needs such as food and beverages, clothes and shoes, health and hygiene etc. These expenditures are divided into gender- and age-related expenses. The household-specific consumption expenditures vary by household size. This includes items principally used by all members of the household. The standard budget is used as a method to define a given income level. The total consumption expenditureFootnote 13 appears as the independent variable whereas the income level (as opposed to the income) becomes the dependent variable.

The ambition of the Norwegian standard budget is to define a «reasonable» consumption level. This is neither subsistence nor a luxury level, but a consumption level that satisfies the welfare ambitions in Norwegian society. The budget is not dedicated to distinct social or economic strata, but is—in principle—suitable for the «average» household in Norway. We will not go into details of the theoretical, methodological and practical implications of this work. In this context we are more concerned with the implication of using the standard budget as a method to define a policy oriented welfare level for families with children.

3.4 A Practical Approach

To support the main arguments in this work, we will demonstrate how the standard budget calculates the consumption needs for single parents with children of different ages to illustrate the cost structures of children in different age groups, both in absolute terms (NOK)Footnote 14 and in relative terms. The standard budget calculations will be compared with the modified OECD-scale, and the implications of these two methods in defining the support level for actual lone parent families will be compared.

In Fig. 1 we have presented the cost calculations for four household types; a single adult, an adult with a 2year old child, and with a 6 and 12years old child respectively. The consumption expenditure for a 2year old child is NOK 2150Footnote 15 per month, while the monthly expenditure for a child of 12years is NOK 3615Footnote 16. This means that the standard budget estimates that it cost 68% more to satisfy the consumption needs for a 12year old child than for a 2years old child. Compared to the modified OECD scale that assumes that all children under 14 have an identical cost profile, the standard budget calculations for the same age groups vary substantially.

Fig. 1
figure 1

Standard budget for single household, single parent with a 2, 6 and 12 year old child. NOK per month

Full size image

In Fig. 2 this is translated into equivalence scales.

Fig. 2
figure 2

Standard budget for single household, single parent with a 2, 6 and 12 year old child. Cost scales

Full size image

The cost scale for the four household types, with the single adult as a reference point, shows that the standard budget on average is in line with the OECD scale that states that a child under 14years has a consumption need that is 30% of the first adultFootnote 17. But we also recognise the relative differences between the age groups. In Fig. 2 a 6year old child has a cost profile equal to the modified OECD-scale. If all children under 14years received NOK 2 275Footnote 18 in social benefit each month to compensate for the cost of children (as is the standard budgets’ cost estimate for a 6year old child), a single household with a 2year old child would receive NOK 625Footnote 19 more than the consumption “need”. A single household with a child of 12years would receive NOK 840Footnote 20 less than the consumption need estimate. This amounts to a yearly “profit” of 7,500Footnote 21 NOK for the former and a “deficit” of 10,080Footnote 22 NOK for the latter, a substantial amount of money for poor people.

4 Discussion

How do we interpret these figures? Do these calculations demonstrate that the standard budget is right and the equivalence scales approach is wrong? Yes and no. The answer must be negative in the sense that there is no actual consensus that the budget figures actually reflect children’s consumption need. On the other hand—and in opposition to the equivalence scales—the empirical and normative aspects of the budget are spelled out explicitly and give the opportunity for a very specific discussion of the rationale behind the calculations. As in all efforts to define welfare levels, these discussions must be open to take factual and normative aspects and implications into consideration. As it comes to the equivalence scales the cost of children is a fixed figure relative to the first adults’ income. This relation is constant for all OECD countries and for all levels of income and there is no explicit normative criteria open for discussion in spite of the substantial consequences in practical use. These objections are not only valid for equivalence scales, but more principally against income approach to welfare levels. The income approach is—as the standard budget—an expert approach. The principal difference is that former avoid making normative decisions in the open while the budget approach has build normative reasoning into its methodological apparatus.

The Standard Budget calculations demonstrate that there are significant differences between measures of levels of living constructed as monitoring devices—development of inequality, international comparisons etc.—and measures used as input in social policy. If, for example, these two methods are used to decide how much money poor families need in order to be out of poverty, the consequences—both for the recipients and the legitimacy of the policy—would be very different. Our hypothesis is that an absolute, substantial approach that explicitly incorporates vital normative arguments combined with empirical knowledge of living standards and consumption needs is more legitimate and accurate than an income approach that hides normative implications in statistical measures and takes preferences for given.