Overview: Social Policies and Child Well-Being

Abstract

This chapter reviews the help that the state provides for parents in a comparative (mainly European Union) perspective. It identifies the different ways that help can be assessed, including comparisons of expenditure on families as a proportion of GDP; the extent to which child poverty is reduced by transfers; the analysis of those transfers using micro data; the use of model families to explore the level and structure; and the package of financial transfers in respect of children. This chapter then explores the association between these efforts on behalf of children and their well-being. There is a clear association between the generosity of state effort on behalf of children and the more objective domains of well-being – material, health, education, housing, and behavior, as well as with overall well-being. However spending on families with children as a proportion of GDP is not significantly associated with subjective well-being, though subjective well-being is associated with the other domains and with income poverty and inequality. It is very likely that the impact of social policy on subjective well-being is mainly indirect, though one can think of measure like anti-bullying strategies which are known to have a direct effect on subjective well-being. Governments can and do support parents in the burdens of child rearing and those countries which make more effort tend to have children with higher levels of well-being.

1 Introduction

The well-being of children is mainly the responsibility of parents. However, in rich countries and increasingly in poor countries the governments help parents. The normative framework for that help is now the UN Charter on the Rights of the Child. Governments are the duty holders under the Charter and they can and are held to account for their performance every few years. Some countries affirm obligations to children in their national constitutions. However, public policy for children has a longer history beginning with legislation to control child labor, the provision of primary education, public health measures, poor relief, and the protection of orphans and vulnerable children in the nineteenth and early twentieth century. Later in the twentieth century, governments began to contribute to the costs of child rearing.

Early public interventions to improve child well-being were probably motivated by religious or philanthropic ideals. However it has come to be accepted that the case for not leaving responsibility for children solely in the hands of parents is much more to do with public social and economic interests. Public policy has been developed to enable parents to reconcile work and family life – to aid economic success. Public policy invests in children to ensure the best possible human capital – a healthy and well-educated population is an essential component of economic success. Children raised in poverty are likely to be less successful in adulthood. Poverty is associated with a host of social problems. Preventing or reducing child poverty is a way of tackling those problems – ill health, low educational attainment, squalor, crime, and unhappiness.

The most radical public involvement in the lives of children is when government takes over responsibility by taking children into care. Children whose parents are struggling to care may need the extra support of child protection services. The direct provision or regulation of preschool childcare has been a developing role for governments. These issues are dealt with in separate chapters in this section of this handbook. This chapter will focus on three elements of public policy for child well-being. First, this chapter will review variation in the efforts that governments make on behalf of children and how it has changed over time, especially as a result of the global economic crisis since 2008. Second, it will compare the level and structure of financial support provided by governments. Third, it will review the outcomes of this support: What evidence is there that this effort results in improved child well-being? The evidence presented in this chapter will be mainly comparative, but inevitably the comparative data is mainly available for rich countries.

2 Government Effort on Behalf of Children

Government effort on behalf of children can be analyzed and compared in three main ways. By analyzing public social expenditure on children; by assessing the extent to which that expenditure reduces child poverty; and by estimating the extra resources that families with children get, compared to single and childless households.

2.1 Social Expenditure on Children

The largest single component of public expenditure on children goes in education. In 2007, the OECD countries spent on average 4.6 % of GDP on primary, secondary, and tertiary education. This varied from 7.0 % in Denmark to 3.1 % in Japan. Norway spent $10,146 on each primary school child and $14,121 on each secondary school child compared with Mexico spending $1,928 per primary child and $1,955 per secondary child (OECD 2012).

It is possible that the second largest component of government effort on behalf of children is child health spending, but it is difficult to separate health spending on children from spending on adults in national accounts and the OECD does not publish comparable data on health spending on children.

The third component and the one that will be the focus here is expenditure on family benefits and services. Family benefits and services include family allowances, maternity and parental leave, “other” cash benefits, day care or home help services, and “other” benefits in kind. On average, in 2007, in the OECD 2.2 % of GDP was spent on family benefits and services and tax breaks for children. This varied between 2.7 % in France and 0.66 % in Korea (OECD 2012). Table 103.1 shows spending per child as a proportion of GDP per capita varies considerably between countries with the USA at the bottom of the league table in 2007 with 3.2 % and Sweden at the top with 20.8 %. It also shows that expenditure on family benefits and services rose in almost all OECD countries between 1980 and 2007 (except the USA and Sweden and Denmark declined since 1990) and this is despite the pressures of population aging. It has also increased in the other countries for which there is data for only some of the period including the Czech Republic, Hungary, Iceland, South Korea, Mexico, Poland, and Slovakia (Bradshaw and Holmes 2013).

Table 103.1 Expenditure per child as a proportion of expenditure per capita GDP in $ purchasing power parities

This data is the latest available at the time of writing, but it is now rather old. Since 2007, there has been a global financial crisis and economic recession with many countries struggling to reduce deficits in their public accounts. Governments may not be able to control the extent to which economies are hit by crisis or how this affects households, but they can control how austerity measures are distributed. They can decide how and whether to cut benefits or wages; or to increase direct or indirect taxes; or to cut of freeze public sector earnings. These decisions will influence the distributional consequences of the responses to the crisis. There is a EUROMOD analysis that evaluates the distributional impact of the tax and benefit measures taken in response to the crisis in six countries (Callan et al. 2011). It shows the contrasting outcomes for pensioners and children. Households with children lost much more than the elderly in Ireland, Spain, and Estonia and if they were poor in Portugal or rich in the UK. In contrast, in Greece the elderly lost more than children.

2.2 Poverty Reduction

A measure of the how well countries are doing on behalf of their children is their child poverty rate and a measure of the success of their efforts is how much they reduce child poverty by transfers (Slater 2011). Figure 103.1 is an example of this kind of analysis for the European Union (there is similar analysis for OECD by UNICEF (2012)). It compares the proportion of children living in households with equivalent income less than 60 % of the median. It shows what the child poverty rate would be if there were no transfers and families had to rely just on market income. It also shows how that poverty rate is reduced by the financial help that governments give – the countries are ranked by the extent to which they reduce pre-transfer child poverty that countries make on behalf of children is child poverty. The countries that achieve most reduction in their pre-transfer child poverty are Austria, Slovenia, Ireland, and Finland. The countries that achieve least reduction are Greece, Spain, Latvia, and Italy. In Austria 70 % of child poverty is reduced by their transfer system, whereas Greece only achieves a 15 % reduction.

Fig. 103.1

Before and after transfer child poverty rates (60 % median threshold): countries ranked by percentage reduction in child poverty EU SILC 2010 (Source: Own analysis)

Perhaps as important as the reduction in the child poverty rate is reducing the child poverty gap – how far below the poverty threshold children are living. It is arguable whether it is better for a country to have low poverty rates and large poverty gaps or high poverty rates but poor children living only a little below the poverty line.

Figure 103.2 compares the average child poverty gaps. Before transfers the gaps are largest in Ireland and the UK but also over 50 % in Belgium, Germany, and Denmark. Negative values for post-transfer gaps indicate that for households with children classified as poor before transfers are included, post-transfer incomes now exceed the poverty line on average, not for every household, but on average. In other words, the poverty gap has been completely closed for many households by taking all transfers into account – they have been lifted above the poverty threshold and the poverty rate has fallen. A positive average gap remains in Bulgaria, Spain, Greece, Italy, Latvia, Poland, Portugal, and Romania. These are the countries that achieve least reduction in their poverty gaps (plotted on the right-hand axis). Slovenia, Hungary, Cyprus, and Iceland achieve the largest percentage reduction in their poverty gaps on average.

Fig. 103.2

All families with children: pre-transfer and post-transfer percentage poverty gaps. Percentage reduction in average poverty gaps on right-hand axis

What we do not know from this is how the countries achieve their child poverty reduction – which policies are more or less effective?

One approach to exploring this has been to compare the level and structure of the child benefit package using model family methods. This has been an approach used in Marx and Nelson (2012) and some of that analysis is reviewed below.

Another approach has been to model the odds of being poor using micro data and then introduce an indicator of the level of social assistance income and observe how it changes the odds. Nelson (2012) pioneered this approach in an analysis of social assistance and it has also been applied to benefit packages for single adults with children (Chzen and Bradshaw 2012).

Yet another approach is to use micro data and decompose the different elements that contribute to the pre- and post-transfer income distribution. In a recent example of this approach, Wang et al (2012) used Luxembourg Income Study data to assess the redistributive effects of benefits and taxes. The problem here is that the benefit element remains a black box – it is impossible to tell which of the various benefits that are included in the package are doing the heavy lifting. Recently, Van Lancker et al (2012) made an advance by identifying the contribution of child benefits to poverty reduction using EU Survey of Incomes and Living Conditions (SILC) 2008. They estimated child poverty rates before and after child benefits and presented the absolute and relative poverty reduction achieved. They found reductions in poverty rates for two adults with children as high as 52 % in Austria and for single mothers 48 % in Ireland.

The data in Table 103.2 takes that approach slightly further by including more elements of the benefit package for children. It is derived from an analysis of EU SILC data for 2010. The first column gives the at-risk of child poverty rate after transfers. The next five columns give the percentage reduction in poverty achieved by each type of benefit. It is effectively an estimate of how much higher the poverty rate would be if that benefit did not exist. So, for example, in Austria child benefits achieve a 62 % reduction in poverty and housing benefits a 7 % reduction. In most countries, child benefits achieve the largest reduction. But in Bulgaria, Greece, and Poland it is old-age benefits – these may be children living with grandparents possibly in multi-generation households. In Denmark and Spain, it is out-of-work benefits that contribute most to poverty reduction. In Denmark, this is because most poor children are in out of work households. In Spain, it is because there are no child benefits.

Table 103.2 All households with children: percentage reduction in child poverty achieved by each element of the benefit package

What is clear from this analysis is that government support for children does not just come in the form of child benefits but is made up by a package of provision.

2.3 Social Protection for Children

Social protection actually starts with the regulation of employment and earnings. The income of families with children is primarily determined by what parents can earn. In most rich countries (The Nordic countries (Denmark, Finland, Norway, Sweden) do not have a minimum wage. Neither do Germany or Switzerland. In Italy, the minimum wage is sectoral (assumed leather workers).), the state ensures that employed parents are paid a minimum wage and that women are treated fairly through equal pay legislation.

Then the state contributes resources to help families with children. There are two main elements to the social protection package for families with children. First, there are the arrangements that exist for families with children where the parent(s) are not employed. In many countries, these will be contributory insurance unemployment or disability benefits. However in almost all rich countries, there is a minimum income or social assistance scheme which provides a floor to living standards. Currently, the most up-to-date data on this floor is the OECD Benefits and Wages database for 2010. The package available to families with children is made up of a combination of the following:

• Means-tested social assistance is paid in all countries except Greece and Italy.

• Family benefits are paid in all countries except Italy, Korea, and Switzerland.

• Housing benefits are paid to tenants in all countries except Belgium, Canada, Greece, Italy, Portugal, Spain, the USA, Bulgaria, Lithuania, and Romania. (Housing benefits are difficult to handle in comparative research. Some countries which do not have separate housing benefits include a standard amount for rent in their social assistance. OECD Benefits and Wages assumes rent is 20 % of average earnings which is likely to be unreasonably high for social assistance recipients.)

• Then there are tax benefits in Australia, Austria, Belgium, and Canada.

• Then the package is subject to deductions for income tax or social security contributions in Denmark, Luxembourg, New Zealand, Iceland, and Israel.

The level of the resultant package (excluding housing benefit) for a couple with two children in 2010 is compared in Fig. 103.3. The amount varies from (leaving aside Greece) $2,909 per year in Romania to nearly $37,000 per year in Luxembourg or from 20 % of the average wage in Romania to 75 % of the average wage in Ireland.

Fig. 103.3

Social assistance for a couple plus two children in purchasing power $ per year and as a percentage of the average wage, 2010 (Source: Own analysis of OECD Benefits and Wages data base 2010)

With the possible exception of Lithuania, this package is not likely to lift any family above the 60 % of median poverty threshold. However, it contributes to closing the poverty gap.

The second element for families with children is the package paid for workers. This consists of the following elements:

Universal child benefits: They exist in all countries in the EU countries except Czech Republic, Italy, Lithuania, Poland, Portugal, Spain, and Slovenia. Among the OECD non-EU countries, Canada, Israel, and Japan have universal child benefits. Argentina is the only middle-income country with a universal child benefit although the child support grant in South Africa reaches 80 % of children.

Means-tested child benefits: All the countries that do not have universal child benefits in the EU or OECD have means-tested child benefits payable to low-income families. In addition, France, Ireland, the Netherlands, Belgium, and Romania have means-tested benefits as a supplement to their universal systems. In some cases, these benefits go quite a way up the earnings distribution. So in the EU, couples with two children and one earner on the average wage would still receive some means-tested child benefit in Bulgaria, France, Italy, Lithuania, the Netherlands, Portugal, Romania, and Slovenia.

Tax benefits: A large number of rich countries also have child tax benefits. In some countries, these are child tax allowances which reduce the tax payable if there are children in the household and are of most benefit to the higher paid and no benefit at all for those earning below the tax threshold. EU countries with this kind of scheme include Greece, Latvia, Poland, Portugal, Slovenia, Slovakia, and Spain. The other kind is child tax credits – cash supplements which may reduce as earnings increase. In the EU, these include Austria, Belgium, Czech Republic, Estonia, France, Germany, Italy, Lithuania, Romania, and the UK. They also exist in the USA, Canada, Australia, New Zealand, and Japan. Tax benefits are quite difficult to classify. For example, the child and earned income tax credits in the UK are administered by the tax authorities but have the characteristics, and appear in the national accounts, as means-tested benefits. The tax credits in Germany, Lithuania, and the Czech Republic have more the character of universal child benefits as they are not exhausted as income increases.

Housing benefits: A number of countries mitigate the housing costs of tenants with low earnings through housing allowances which are more generous if children are present. In the EU, these include Austria, Czech Republic, Finland, France, Greece, Hungary, Italy, the Netherlands, Poland, and Slovenia.

Social assistance: In addition, some countries pay social assistance to low-paid workers. In the EU, these include Bulgaria, Czech Republic, Finland, Lithuania, Luxembourg, Portugal, and Slovenia.

There are other potentially valuable benefits that are not included in the modeled package – including the mitigation of childcare costs (see below), free or subsidized school meals, and in the USA food stamps (SNAP) is an important part of the package.

Figure 103.4 derived from MIPI data (Van Mechelen and Bradshaw 2012) compares the structure and level of the child benefit package for a low-paid family in the EU (plus Norway) and three US states (Nebraska, New Jersey, and Texas). This is estimated by taking the net incomes of a couple with two children from the net income from a childless couple on the same earnings. It is thus a measure of the contribution of the state to child rearing. It illustrates the variation in both the level and the complexity of the structure of the child benefit package.

Fig. 103.4

The structure of the child benefit package for a couple plus two, in purchasing power standards, June 2009. Euros per month. Single earner – minimum. Based on the composition of child benefit packages a model family with two children (aged 7 and 14): a single earner couple – a single earner couple – minimum wage (or half average wage for countries without a statutory minimum wage: DE, DK, FI, IT, SE). UCB Universal cash benefits, IRCB Income-related cash benefits, CTB Child tax benefit, HA Housing allowance, SA Social assistance, Other Local property and other non-income taxes + social security contributions. Average earners in Greece were assumed to be civil servants, and therefore receive civil servants’ family allowance. Minimum wage earners were assumed to receive OAED child benefit

Figure 103.5 shows the same observation for the same family but this time for an earner receiving the average wage (no data for Luxembourg).

Fig. 103.5

The structure of the child benefit package for a couple plus two¹, in purchasing power standards, June 2009. Euros per month. Single earner – average

Comparing these two figures, it is possible to estimate a degree of progressivity of the child benefit package – to what extent is it concentrated on the low paid. In Latvia, Estonia, Spain, Slovakia, and Greece, the package is regressive – more generous to the higher earning family – this is the result of child tax allowances. In Germany, Denmark, Norway, Sweden, and Belgium, the package is effectively neutral. Then the other countries are more progressive with the package for the low-paid family worth more than three times that for the average paid family in the Czech Republic, Lithuania, Portugal, and New Jersey.

There is considerable debate in the literature about whether or not child benefit packages that consist of mainly income-related benefits provide better minimum income protection for low-paid workers than ones that are more universal (Kenworthy 2011). Although targeted systems may in theory be more generous to low-income families, they may be quite ineffective with regard to poverty alleviation due to take-up problems and labor market disincentives (Deacon and Bradshaw 1983). In practice, countries with the most generous child benefit packages tend to combine universal benefits with income-related cash benefits, housing allowances, or supplementary benefits from social assistance.

Figure 103.6 plots the degree of progressivity against the percentage reduction in child poverty achieved by transfers. It can be seen that there is no clear relationship. There is certainly no evidence here that highly progressive child benefit packages are inevitably more likely to be effective in reducing poverty.

Fig. 103.6

Degree of progressivity by percentage reduction in child poverty (Source: Own analysis)

Some of these problems of means-testing are illustrated using Fig. 103.7 which depicts the child benefit package paid to a couple with two children in the UK. (In principle, similar figures could be produced for other countries using the OECD tax benefit database, but in practice it is not yet flexible enough to control for the interaction of benefits.) When there are no earnings (nil hours worked), the family will receive child benefit, child tax credit, social assistance (Job Seekers’ Allowance), housing benefit, and council tax benefit totaling about £320 per week. When there is one earner on the minimum wage net earnings begin to replace social assistance, until at 16 h a week in the UK system the family shifts to the in-work benefit system. There are still the universal child benefit, child tax credit, but now earned income tax credit, net earnings, housing benefit, and council tax benefit. There are three things to observe. First, the distribution is very flat – as earnings increase, earned income tax credit, housing benefit, and council tax benefit are reduced and the earner pays more income tax and social insurance contributions. As a result, the family faces a very high marginal tax rate. Second, as a consequence the net income for working full time is only £70–£80 higher than not working at all. Third, the family does not reach the poverty threshold even working full time, and this is the reason why over half of poor children in the UK have a parent in employment.

Fig. 103.7

UK: Net disposable income for a couple plus two children before housing costs by hours supplied at the minimum wage from April 2009. Rent = £60 a week, Council Tax = £18.00 a week

Arguments in favor of universal benefits and against means-tested benefits have recently been reviewed (Bradshaw 2012). Briefly, the arguments against means test are as follows:

• They tend to be low benefits because they lack the political support of universal benefits.

• They tend to be too targeted – concentrating on the very poor, whereas many more families with children are struggling financially.

• As a result, they are not much more effective than universal benefits in reducing poverty rates or closing poverty gaps.

• Not only is coverage poor but means-tested benefits suffer from low take-up as a result of ignorance of the existence of the benefit or that one might be eligible, or stigma associated with receiving it. If only a small minority are eligible, then few realize they are eligible.

• Then there are the complexities in demonstrating eligibility, the need to produce documents.

• Means tests are usually jointly assessed and with more fluid families this now makes the resulting benefits much less stable and regular in terms of income for children, as they will get disrupted each time there are changes in circumstances.

• The high marginal tax rates associated with means-tested benefits produce unfortunate behavioral effects. They reduce incentives to enter and stay in employment, increase earnings more, and to save.

These problems are perhaps best illustrated by the experience of the countries that once formed the Soviet republics. Before the collapse of the Soviet Union and the other state socialist countries in the CEE/CIS Region, children were the focus of fairly effective social protection policies. Generally there was free health care, free education, a system of state preschool child nurseries, full employment, and universal child cash benefits. Of course it was not perfect, but it provided a level of security against extreme poverty. These systems were largely swept away during transition. In a number of countries, user charges were introduced for health and for education with a rapid growth of informal payments. State nurseries were closed and parents needing preschool provision were obliged to pay for it. Unemployment grew and while some countries (re)introduced contributory unemployment insurance, this only protected those in the formal economy. In many countries in the region, parents were forced to travel abroad to find work. Family incomes fell with a shift from two-earner to one-earner and no-earner households. Cash transfers to families with children became means-tested. In most countries, the only cash benefit available to families with children is through so-called targeted social assistance (TSAs). The TSAs in the region share common characteristics. They are commonly payable to households on the basis of a quasi means test, often of great complexity. The means test is designed to exclude households with any resources from earnings, home production, or gifts from relative or friends. This is despite the fact that the majority of children in poverty are living in households with some employment income. The level of the benefit varies from country to country but it is very low. Often it is below the food poverty threshold – not large enough to live on without other resources, which they are not supposed to have. Despite this much effort is made to improve targeting by including the large numbers of poor families with children who do not apply/receive the benefit and exclude households who are receiving but are formally not eligible, although they may be very poor by any standard.

These schemes were commonly introduced on the advice of and with technical and sometimes financial support of the World Bank. The Bank has also been the dominant influence on poverty measurement as funder of household consumption surveys, and then leading analysts using their $x per day consumption thresholds. They are pretty useless systems, incapable of transferring the benefits of growth to families with children and leaving these countries with very high child poverty rates (Bradshaw et al. 2012).

3 Policies and Child Well-Being

Do these policies have an impact on child well-being? There can be no doubt that government spending on education and health for children has an impact on educational attainment and participation and on child health. There will of course be debates in the educational field about the merits of investing in early years, primary, secondary, or tertiary on educational outcomes. Similarly, in health there will be arguments about the merits of spending on prevention or treatment.

3.1 Cash Versus Care

It is also clear that government transfers for children have a big impact on child poverty – the evidence has been shown earlier in this chapter. However, in relation to family benefits, there is a debate in progress about whether it is better to spend money on cash transfer or services in kind. In the UK, the Field Review (2010) actually argued that investments in services should be funded by cutting child and family benefits. The Chancellor immediately responded by reneging on his commitment to uprate child tax credits above the rate of inflation and instead to invest in early years for some deprived 2 year olds.

The claim that spending on services is better than spending on cash benefits may be influenced by OECD who claim on their website (http://www.oecd.org/edu/earlychildhood.) childcare…“brings a wide range of benefits, including social and economic benefits ; better child well-being and learning outcomes ; more equitable outcomes and reduction of poverty ; increased intergenerational social mobility ; higher female labour market participation and gender equality ; increased fertility rates ; and better social and economic development for society at large.” But this is an association not a cause. These countries have an egalitarian income distribution, high levels of parental labor market participation, high wages and, yes, heavy investment in good quality childcare. They also start with comparatively low pre-transfer child poverty rates. Spending on cash benefits fell in all the Nordic countries in the 2000s and their child poverty rates increased.

The association between spending on services and child poverty is shown in Fig. 103.8. There is a fairly weak association – thanks mainly to the Nordic countries. However, there is no association between the proportion of family spending spent on services and child poverty rates or gaps.

Fig. 103.8

Child poverty rate by spending on family services as percentage of GDP

There is a much stronger association between spending on cash benefits and tax breaks and child poverty rates than there is with spending on services. In EU countries, there is a stronger association between child poverty gaps than child poverty rates and spending on cash benefits and tax breaks. There is an even stronger association between spending on cash benefits and tax breaks and the reduction in child poverty achieved by transfers. See Fig. 103.9.

Fig. 103.9

Percentage reduction in child poverty by percentage of GDP spent on cash benefits and tax breaks. EU countries

Actually it is the level of total spending on families – cash benefits + services + tax breaks – that is most closely associated with child poverty. The lesson of this observation is that we need to invest in children in all sorts of different ways. Spending on childcare probably helps to increase maternal employment, enhances gender equality, and possibly also has beneficial child development outcomes. But it is probably not the best way to tackle child poverty and income inequalities. Indeed recent analysis (Van Lancker 2012) of EU SILC data suggests that in many countries childcare for children aged 2 or less reaches the rich better than the poor. It certainly does not tackle the poverty of older children – except possibly in the long term.

3.2 The South African Child Support Grant and Child Well-Being

South Africa is the only country in Africa that has a cash grant for all poor children. There are a few other countries that have cash payments for orphans and vulnerable children (OVC), and among these Namibia is planning to develop their OVC grants toward the more inclusive South African model. Until Argentina introduced a universal family allowance, South Africa was the only middle-income country in the world with a child cash payment of this type. The child support grant (CSG) introduced in 1998 is a means-tested cash benefit and approximately 80 % of children are eligible. It is paid for children aged 0–17 inclusive who live in low-income families (this includes South African citizens, permanent residents, and refugees). The amount is R280 per month for the child (£22.40 per month on current exchange rate). The means test is R33, 600 per year if the primary caregiver is single. If the primary caregiver is married, the means test is R67, 000 per year. There is no asset test. There is no social assistance for unemployed adults in South Africa, so adults without work who have children often find that they also have to rely on the child support grant. The CSG has been subject to quite extensive evaluation. The most recent evaluation (DSD, SASSA and UNICEF 2012) found that child support grant has improved height for age scores and thus cognitive development with long-term benefits. It has also improved maths ability tests and scores on reading and vocabulary for 10-year-olds. Girls’ attainment at age 6 was improved by a quarter, mainly by reducing delays in entering school. “In these ways the Child Support Grant promotes human capital development, improves gender outcomes and helps to reduce the historical legacy of inequality” p iv. The child support grant has also been shown to improve health and reduce the likelihood of illness – boys enrolled early on the child support grant had a 21 % likelihood of being ill compared to 30 % if enrolled late. For adolescents, it has been shown to reduce school absences and reduced the likelihood of working outside the household. It also reduces risk behavior including sexual activity, pregnancy, alcohol use, drug use, criminal activity, and gang membership.

These remarkable improvements in well-being associated with cash grants is one reason as to why UNICEF (Winder and Yablonski 2012) has developed a social protection strategic framework which advocates the use of child cash grants in developing countries and in June 2012, member countries of the International Labour Organisation (2012) adopted the Recommendation Concerning National Floors of Social Protection, which states that member countries should “Establish as quickly as possible and maintain their social protection floors comprising basic social security guarantees. The guarantees should ensure at a minimum that, over the life cycle, all in need have access to essential health care and to basic income security which together secure effective access to goods and services defined as necessary at the national level.”

4 Subjective Well-Being

So there is not much doubt that public social policies for children improve their material circumstances and their health and education. The association between policy and objective well-being is pretty clear. But what about subjective well-being, what children feel about their lives and their relationships?

Happiness has been receiving more attention recently. The observation of the paradox that getting richer does not lead to greater happiness (after a certain level) resulted in the initiative by President Sarkozy of France to set up the Stiglitz Commission (2009). In its report, the Commission urged countries (and the OECD) to collect data on subjective well-being. The OECD began to publish the How’s Life index (http://www.oecd.org/statistics/howslifemeasuringwell-being.htm) and a number of countries have begun to collect data on subjective well-being – for example, the UK government has asked the Office of National Statistics to incorporate questions on happiness of both adults and children in national surveys. This is evidence that governments are concerned about the impact of policy on children’s happiness. Some government policy may actually harm children. Successive PISA surveys have demonstrated that Finland is an outlier – best of all countries in educational attainment. But in the Health Behavior of School Aged Children (HBSC), Finland has one of the lowest proportions of children liking school a lot. UNICEF UK (Ipsos-MORI 2011) launched a comparative qualitative study of subjective well-being of children in Spain, Sweden, and the UK in an attempt to explain the results of RC7. It concluded that UK children did worse because of long parental working hours and high levels of materialism. Materialism may not be blamed on government policy but encouraging both parents to work and then failing to provide adequate policies to reconcile work and family life are certainly policy.

There is certainly national evidence (e.g., the Children’s Society 2012) that shows that children are happier if they live in decent houses, in safe neighborhoods, are not bullied, enjoy and achieve in schools, and are not materially deprived. These can all be influenced by policy. Family and other relationships may matter more than these things and they may not be directly amenable to policy. But indirectly they can be – by for example – reducing the burdens of poverty and inequality on parents, treating parental depression, and providing family friendly services.

There are three types of arguments that might suggest that subjective well-being may not be amenable to public policy.

Some argue that subjective measures are (well) subjective – that asking children structured questions in sample surveys about what they think and feel is somehow less valid or reliable than asking the children questions in the same survey about whether they smoke or drink alcohol. Some may think that subjective well-being is a transient mood and therefore not reliable. Alternatively they may argue that it is culturally determined – that children in, for example, France have a different understanding of “satisfied with life” than they do in, say, Korea. Alternatively it may be argued that the real meanings are “lost in translation,” for example, “the best possible life for you” may mean different things in Japanese to what it means in Italian. There is no real evidence for any of these anxieties, but they are often powerfully held views.

Another argument is that expressed views about health or life satisfaction may be a function of false expectations or adaptive preferences. Very deprived children may say that they are very satisfied with life because they know no better, or they have become reconciled to their lot. There is certainly some evidence in poverty studies (Ridge 2002) of poor children not complaining to their parents in order to protect them from guilt. An example of false expectations in the other direction would be a girl being dissatisfied with her body or her clothing because she does not look like models she sees in the media. There is no systematic evidence of false expectations or adaptive preferences in subjective well-being. No doubt some exists. But nevertheless children’s views are their views.

Then there is the argument that subjective well-being is related to personality, personality is formed by nature and nurture, and nature and nurture cannot really be influenced by public policy. There may be evidence that subjective well-being is related to personality but the problem with this is that the way we measure personality is not independent of subjective well-being.

Then there is the argument of Cummins (2010) that happiness is the result of genetically determined homeostatic adaptation. Over the millennia, humans who have survived more successfully have been those who have had more capacity to adapt to their environment and the shocks of life. Humans, including children, have a natural resilience to bounce back to a predetermined happy state. This may explain why it is hard to explain variations in subjective well-being in terms social structural characteristics or life events – because people have all bounced back. However, the empirical evidence is that not everyone has bounced back to the same level. There is a tail of low subjective well-being and also international variation. What explains it?

UNICEF (2013) created a multi-component domain of subjective well-being derived from HBSC data. It is a composite of four components containing eight indicators summarized in Table 103.3. Each indicator score contributes equally to the component score and the subjective well-being score is an average of the z scores of the four components.

Table 103.3 Subjective well-being

Figure 103.10 gives the rank order of countries on subjective well-being.

Fig. 103.10

Subjective well-being rank orders z scores. Mean = 100

4.1 What Explains This Variation?

How are the subjective components and the overall subjective well-being domain related to the other more objective domains of well-being? Table 103.4 summarizes the associations. Overall subjective well-being is associated with all the objective domains. Overall well-being (excluding subjective) is associated with all the components of subjective well-being except educational well-being. In fact, educational well-being is not associated with any of the objective domains including education. Material well-being is associated with all the subjective components. The health and safety and the housing environment domains are associated with life satisfaction and health well-being but not family relations and education. Behavior is associated with life satisfaction and family relations.

Table 103.4 Correlation between the other domains of well-being, overall subjective well-being, and the subjective well-being components

Figure 103.11 shows the association between overall well-being excluding subjective well-being and overall subjective well-being. Romania, Latvia, and the Netherlands have higher subjective well-being than you would expect given their overall objective well-being. In contrast, the USA, Italy, France, Canada, Poland, and Luxembourg all have lower subjective well-being than you would expect given their objective well-being.

Fig. 103.11

Association between overall well-being excluding subjective well-being and overall subjective well-being

What evidence is there that social structural factors are associated with overall subjective well-being? Table 103.5 summarizes some relationships. There is no (significant) relationship between the wealth of a country and the overall subjective well-being. There is an association between inequality, relative poverty and deprivation, and subjective well-being. Interestingly there is a stronger relationship between inequality and overall subjective well-being than with any of the other objective domains. Perhaps most interesting for the purposes of this chapter is that there is no significant association between the proportion of GDP spent on families and subjective well-being.

Table 103.5 Association between structural factors and subjective well-being

The actual relationship is detailed in Fig. 103.12.

Fig. 103.12

Association between subjective well-being and % GDP spent on families