Spending Better for Children through Social Policy

Report

5 October 2026

3. The value for money of social policies in reducing child poverty

Copy link to 3. The value for money of social policies in reducing child poverty

Abstract

This chapter assesses the “value for money” of social policies in reducing child poverty and material deprivation. It examines how spending across policy areas and specific policy design features affects the effectiveness of social expenditure and influences future spending pressures, while highlighting complementarities between different types of expenditures.

The chapter first reviews key trends in child income poverty and material deprivation. It shows that poverty risks remain particularly high among children living in jobless households, single‑parent families and large families, while also highlighting the prevalence of in-work poverty. It further stresses that material deprivation affects a significant share of children living in households with incomes above standard poverty thresholds.

The analysis finds that pro‑employment policies and in-kind support, such as health, education and childcare services, are particularly effective in reducing child poverty and severe material deprivation while helping to contain growth in social spending. By contrast, income support policies tend to be impactful in mitigating child poverty close to the poverty threshold but less effective in addressing deeper forms of poverty and material deprivation and are associated with stronger pressures on public spending.

It concludes by discussing policy challenges in strengthening the effectiveness of social policies in reducing multiple dimensions of child poverty while keeping social spending sustainable.

Growing up in poverty can have far-reaching consequences for children’s well-being and development. Beyond the immediate hardship associated with limited material resources, poverty can affect children’s emotional well-being, living conditions and opportunities to develop well and reach their potential. A large body of research shows that childhood poverty is associated with poorer developmental outcomes, including lower educational achievement, worse physical and mental health, and weaker labour market prospects in adulthood. Children who grow up in low-income households are more likely to leave school with lower qualifications, experience poorer health over the life course, and face lower earnings and less stable employment as adults (Duncan et al., 1998[1]; National Academies of Sciences, Engineering and Medicine, 2019[2]; Clarke et al., 2022[3]). In this way, childhood poverty contributes to the intergenerational transmission of economic and social disadvantage, limiting opportunities for upward mobility and perpetuating inequalities across generations (Corak, 2013[4]; OECD, 2018[5]; De Schutter et al., 2023[6]).

Social policies play a key role in tackling child poverty and limiting its long-term consequences. They can help prevent poverty by supporting parents’ access to employment, strengthening work incentives and providing opportunities for low-earning parents to develop their skills and improve their labour market prospects. At the same time, social policies can mitigate the damaging effects of poverty on children’s well-being and development. Income support through cash transfers and tax reductions can help lift families with children above the poverty threshold or reduce the depth of poverty when earnings are insufficient. In addition, in-kind support – such as access to early childhood education and care, health services, housing assistance and other family support services – can both help prevent poverty and reduce its impact on children’s living conditions and developmental outcomes (OECD, 2011[7]; Thévenon et al., 2018[8]). Together, these policies form a central component of governments’ efforts to promote children’s well-being and ensure more equal opportunities across generations.

Chapter 2 provided insights into recent evolutions in social spending patterns and related policy design features across OECD countries. These developments suggest a growing alignment with a social investment approach, characterised by strengthened activation policies, expanded support for families with children – particularly during the early years – and greater recognition of the benefits of early prevention. They also include an increasing share of in-kind social expenditures, reflecting greater emphasis on services such as childcare, education and health. At the same time, reductions in the degree of income redistribution through cash transfers appear to signal a gradual shift away from welfare state models that rely predominantly on compensatory income support. The present chapter examines whether these developments have begun to deliver the expected outcomes in terms of reducing child poverty while helping to contain pressures on social spending. More broadly, it analyses how different social policy design features influence both the effectiveness of policies in reducing child poverty and their implications for social expenditure trends.

The chapter is structured as follows. The first section provides an overview of key trends in child poverty. The second section describes how selected OECD countries have designed their child poverty strategies within a social investment framework. The third section presents the analytical approach used to assess these relationships, while the following section summarises the main findings (with further details provided in the chapter Annexes). Building on these results, the concluding section discusses the main policy challenges in designing social policies that more effectively reduce child poverty while containing pressures on social spending. The analysis of this chapter does not consider the effects of social spending on child outcomes beyond income poverty and material deprivation – which are examined selectively in the next chapter.

3.1. Child poverty trends

Copy link to 3.1. Child poverty trends

Child poverty is a multifaceted phenomenon, encompassing low income, material deprivation, and financial stress, which may occur individually or in combination (Schenck-Fontaine and Panico, 2019[9]; Cooper and Stewart, 2021[10]). These dimensions often reinforce one another, with cumulative effects on children’s outcomes. Income poverty, for instance, limits parents’ ability to invest in childcare and educational resources, undermining children’s cognitive development. It also increases the likelihood that families live in neighbourhoods with poorer infrastructure, lower-quality schools, fewer services, and greater exposure to crime and environmental hazards (Chetty et al., 2014[11]; Chetty and Hendren, 2018[12]; OECD, 2025[13]). In addition, financial strain associated with low income can generate stress for parents, which may affect parenting practices and children’s outcomes across multiple domains (Masarik and Conger, 2017[14]). Material deprivation may accompany income poverty but can also affect families with income above the poverty line. When low income, material deprivation, and financial stress overlap, children face the greatest risk of behavioural problems (Schenck-Fontaine and Panico, 2019[9]).

Measuring child poverty is not straightforward, as it inevitably relies on conventions that classify some families as “poor” while excluding others whose incomes are only marginally higher but whose living standards and financial difficulties may be very similar. In high-income countries, the most common approach is to estimate children’s “theoretical” income based on parental or household disposable income, typically against a relative poverty threshold that reflects a country’s general living standards. Relative income poverty then can be interpreted as an indicator of poverty risk, or a proxy of as opposed to actual hardship experienced by children living in low-income families. Parents with limited resources often prioritise children’s needs in their spending decisions, which can partly shield children from the full impact of low income on their well-being (Cooper and Stewart, 2017[15]; Page, 2024[16]). At the same time, living on a low income can create insecurity and stress within the household, which may affect parental well-being, parenting, and, in turn, children’s outcomes (Schenck-Fontaine and Panico, 2019[9]).

While income‑based measures of poverty are convenient for international comparison and provide useful information on income inequality and the prevalence of low-income households, it has important limitations as an indicator of poverty (Thévenon et al., 2018[8]; Perry, 2021[17]). Households with similar current incomes can differ significantly in their access to non-income resources (such as savings, assets, or external support) and in their financial obligations (including housing costs, health expenses, or debt), meaning that income alone does not reliably distinguish between those who can maintain an acceptable standard of living and those who cannot. Although measures such as income after housing costs improve accuracy, they do not fully address these shortcomings (Perry, 2021[17]; OECD, 2025[18]). Importantly, while income remains a key determinant of material well-being – particularly for low-income households – it is not, on its own, a sufficient measure for identifying economic hardship.

A growing body of research calls for complementing income‑based measures with additional indicators – such as on material deprivation and financial insecurity – that may more directly capture the material and psychological conditions shaping children’s well-being and development. In line with this, countries have started to develop complementary indicators with the aim to provide a more comprehensive picture of households experiencing poverty (see Box 3.1 for more details on indicators developed by the European Union such as the risk of poverty or social exclusion “AROPE”, which captures all those who are either at risk of poverty, or severely materially and socially deprived, or living in a household with a very low work intensity).

Material deprivation indicators, however, also have limitations. They rely on predefined lists of items and living conditions, which inevitably reflect choices about which aspects of consumption should be considered essential. As a result, they may omit other dimensions that also matter for children’s development (such as for instance personal toys, bicycle or other independent mobility tools, access to safe outdoor play spaces nearby). In contrast, income poverty measures do not depend on the selection of specific items and may therefore capture children who experience financial strain even if they are not deprived according to the items included in deprivation indicators. They are also informative in highlighting the limited capacity of low-income households to plan and invest in their children’s development beyond meeting immediate needs. Taken together, these considerations suggest that using both income poverty and material deprivation indicators provides a more comprehensive picture of child poverty.1

One in eight children is at risk of income poverty

Child poverty is most commonly measured through a relative income poverty measure, which captures the share of children living in households with an equivalised2 disposable income below a defined poverty threshold. By convention, the OECD sets this threshold at 50% of the national median equivalised disposable income.

In 2023, one in eight children under the age of 18 across the OECD (12.5%) were at risk of income poverty – a rate that in many countries exceeds that of the total population (Figure 3.1, Panel A). Child poverty rates are particularly high in countries shown on the right side of Figure 3.1 Panel A, including Chile, Costa Rica, Israel, Spain, the United States, Türkiye and Mexico, where children are disproportionately affected by income poverty. Conversely, child poverty rates are lowest in countries on the left-hand side of the figure, such as in Denmark, Finland, Hungary and Slovenia, where children are less likely than the overall population to live in income‑poor households.

An alternative measure is particularly useful for analysing poverty trends over time against a fixed benchmark: the anchored child poverty rate. This indicator measures the share of children living in households whose real living standard – adjusted for price inflation – remain below the poverty threshold defined at a specific point in the past (here, 2005), usually lower than the poverty line estimated in the current year.3

Figure 3.1, Panel B compares the 2022 relative poverty rate with an anchored poverty rate, using a threshold based on 2005 income levels adjusted for inflation. In most countries, anchored poverty rates are lower than relative poverty rates, indicating that median incomes – and therefore relative poverty lines – have increased over time. As a result, some children counted as relatively poor today have living standards that, in absolute terms, are higher than those of children considered poor in 2005. The figure suggests that, among children in relative income poverty, around two‑thirds (about 8% of the total child population) fall below the inflation-adjusted 2005 threshold, while the remaining one‑third (around 4% of the total child population) are above this lower benchmark.

The relationship between relative and anchored child poverty varies substantially across countries, reflecting differences in income growth and distribution over time. In contrast to most countries, in Costa Rica, Greece, Lithuania and Hungary the anchored poverty rate exceeds the relative poverty rate. This indicates that relative child poverty since 2005 are not driven by improvements in living conditions among poorer households, but rather by falling median living standards, which lower the relative poverty threshold and signal a deterioration in living standards across the lower half of the income distribution.

Family structure also affects child poverty rates, particularly whether a household is headed by one or two parents. Across all countries, single‑parent households face a significantly higher risk of income poverty – on average, they are three times more likely to be at risk compared to two‑parent families (Figure 3.1, Panel C).

Figure 3.1. One in eight children (13%) live in income poverty across the OECD
Copy link to Figure 3.1. One in eight children (13%) live in income poverty across the OECD

Child poverty has fluctuated with the economic cycle over the past 15 years. Both relative and anchored child poverty rates rose sharply in the wake of the 2008 global financial crisis and remained persistently high until the mid‑2010s (Figure 3.2, Panel A). Thereafter, both measures declined, with extensive policy support during the COVID‑19 pandemic helping to cushion the economic impact of lockdowns (Dely, Hyee and Prinz, 2025[20]; Daly and Ryu, 2025[21]). Since 2021, however, child poverty has been on the rise again, reflecting the impact of the cost-of-living crisis and surging energy prices (UNICEF Innocenti, 2023[22]; Hiilamo and Lallukka, 2023[23]).

Trends in child relative poverty differ markedly across countries (Figure 3.2, Panel B). In around two‑thirds of OECD countries, child poverty rates have declined since the aftermath of the financial crisis, with particularly large reductions (of around 7 percentage points (p.p.) or more) observed in Greece, Hungary and Korea. Among countries with data extending back to the mid‑2000s, roughly half recorded decreases in relative child poverty, with the largest declines in Ireland, Latvia and Poland. By contrast, child poverty has risen compared with mid‑2000s levels in Sweden (+4.3 p.p.), Norway (+3 points), the United Kingdom (+2.6 points) and Germany (+2.5 points). Nevertheless, child poverty rates in both Norway and Sweden remain well below the OECD average as of 2022.

Figure 3.2. Child income poverty trends are shaped by economic recessions
Copy link to Figure 3.2. Child income poverty trends are shaped by economic recessions

Employment provides important protection against poverty. Nevertheless, a significant share of families with children experience income poverty despite having at least one working parent. On average, around 10% of working families with children are income poor (Figure 3.3). In-work poverty is particularly prevalent among single‑parent families: on average, about 20% of working single‑parent families live in income poverty, and in some countries like Costa Rica, Spain and the United States this share exceeds one‑third.

Figure 3.3. Around one tenth of families with working adults are income poor
Copy link to Figure 3.3. Around one tenth of families with working adults are income poor

Percentage of income poor families with children with at least one working adult, by partnership status, 2023 or latest available year

The effect of income transfers on child income poverty

Figure 3.4 illustrates the critical role of taxes and transfers in reducing child income poverty across OECD countries. On average, redistribution lowers child poverty rates by around one‑third – equivalent to about 7 p.p. – highlighting its central importance in supporting children in disadvantaged households. The impact is strongest in relative terms in Finland (‑73%), Ireland (‑68%), Poland (‑67%), Mexico (‑63%) and Belgium (‑60%). By contrast, redistribution reduces child poverty by half the OECD average or less in Luxembourg (‑17%), Korea (‑16%), Spain (16%), Costa Rica (‑13%), Israel (‑12%), Türkiye (‑8%) and Greece (‑4.6%). In Japan and Switzerland, income transfers are not linked to reductions in child income poverty.

Figure 3.4. Child income poverty rates fall markedly after taxes and transfers
Copy link to Figure 3.4. Child income poverty rates fall markedly after taxes and transfers

Child relative income poverty rates before and after tax and transfer, 2023 or latest available

Income poverty increases the risk of hardship in many areas of children’s lives

The consequences of income poverty span many dimensions of children’s living conditions and access to basic services. For example, children living in income‑poor households are more likely than others to live in poor-quality housing, have unmet health needs, and lack regular leisure activities (Figure 3.5). At the same time, while low income significantly increases the risk of experiencing these difficulties, a non-negligible share of children who are not income poor also face similar challenges.

Figure 3.5. Children in income‑poor households are more likely to live in poor-quality housing, have unmet health needs, and lack regular leisure activities
Copy link to Figure 3.5. Children in income‑poor households are more likely to live in poor-quality housing, have unmet health needs, and lack regular leisure activities

Child material deprivation

While income poverty is a widely used indicator of economic disadvantage, it does not fully capture the material hardships that children may experience. Families with low income often prioritise spending on their children, meaning that low household income does not automatically translate into child material deprivation (Cooper and Stewart, 2021[10]) and public services like education and care, housing and nutrition programmes may further shield children from the detrimental impacts of low income. Conversely, material deprivation can also affect families whose income is above the poverty line, highlighting that not all children who experience deprivation are classified as income poor (Guio et al., 2018[26]). Examining child material deprivation alongside income poverty therefore provides a more complete picture of children’s living conditions and needs, capturing hardships that income measures alone may miss.

In Europe, several indicators have been developed to measure children’s exposure to material deprivation, each with its own strengths and limitations (Box 3.1). One widely used measure is the material and social deprivation indicator, which identifies individuals experiencing an enforced lack of essential goods, services, or social activities due to financial constraints. The indicator is defined as the share of the population unable to afford at least 5 out of 13 deprivation items, covering core aspects of living standards such as keeping the home adequately warm, affording meals with adequate nutritional content, coping with unexpected expenses, or participating in regular leisure and social activities. Information on the presence of children in the household is then used to estimate the proportion of children living in households exposed to severe material and social deprivation.

Across Europe, around 12% of children under the age of 16 were exposed to material and social deprivation in 2025, reflecting a marked decline since the mid‑2010s (Figure 3.6). Reductions have been particularly large, both in absolute and relative terms, in Latvia (‑69%), Lithuania (‑62%) and Hungary (‑60%), as well as in relative terms in Poland (‑70%), where deprivation levels were already comparatively low in the mid‑2010s.

Figure 3.6. Material and social deprivation affect around one in eight children in European OECD countries
Copy link to Figure 3.6. Material and social deprivation affect around one in eight children in European OECD countries

Percentage of children below 16 years at risk of material and social deprivation, 2025

Box 3.1. The European Indicator on child material deprivation

Copy link to Box 3.1. The European Indicator on child material deprivation

Tackling child poverty and social exclusion is an objective of European Union policies, including the European Child Guarantee. The European Union Social Protection Committee does not have a single and formally approved indicator to measure the prevalence of child poverty and social exclusion. Eurostat has developed a few different indicators to measure child poverty and the risk of social exclusion for children, to which the Social Protection Committee refers to (Eurostat, 2023[27]).

In the European Union, child deprivation is measured by Eurostat in two ways: annually through the Material and Social Deprivation (MSD) rate, and every three years via the child-specific material deprivation indicator.

The MSD rate, part of the EU-SILC data, captures the proportion of the population experiencing an enforced lack of at least 5 out of 13 essential items (6 individual-level, 7 household-level) and is used in the EU’s AROPE measure for monitoring poverty and social exclusion. Items include items at household level: Capacity to face unexpected expenses; Capacity to afford paying for one week annual holiday away from home; Capacity to being confronted with payment arrears (on mortgage or rental payments, utility bills, hire purchase instalments or other loan payments); Capacity to afford a meal with meat, chicken, fish or vegetarian equivalent every second day; Ability to keep home adequately warm; Have access to a car/van for personal use; Replacing worn-out furniture; and items at individual level: Having internet connection; Replacing worn-out clothes by some new ones; Having two pairs of properly fitting shoes (including a pair of all-weather shoes); Spending a small amount of money each week on him/herself; Having regular leisure activities; Getting together with friends/family for a drink/meal at least once a month.

For children under 16, items collected only for adults are estimated: if at least half of the adults in a household lack an item, children are considered deprived of it. A lower weight is applied to adult items to avoid overestimating deprivation among children, and at least three of the household items must be lacking for a child to be considered deprived. The indicator thus reflects the share of children living in materially and socially deprived households.

The child-specific material deprivation rate (CSMDR) is another indicator included in the set of indicators proposed by the Social Protection Committee to monitor social inclusion at EU level. This indicator measures the share of children suffering from an enforced lack of at least three of 17 items,1 including child-specific items relating to food and nutrition, clothing and footwear, social activities, leisure activities, and a few household-level items such as whether the household can afford or not to: avoid arrears; have adequate warmth in home; have (access to) a car for private use; replace worn-out furniture; and have an internet connection at home. The adoption of this indicator in March 2018 follows the work by Guio et al. (2018[26]) which establishes consistently high levels of reliability of this indicator.

1. The complete list of items taken into account to measure children’s exposure to material deprivation include information on household’s inability to provide at least one child in the family with: some new clothes; two pairs of shoes; fresh fruits and vegetables daily; meat, chicken, fish daily; suitable books; outdoor leisure equipment; indoor games; leisure activities; to organise celebrations; invite friends; participate in school trips; and go on holiday; it also includes non-child-specific information on whether the household cannot afford to: avoid arrears; have adequate warmth in home; have (access to) a car; replace worn-out furniture; access to Internet.

Source: Eurostat (n.d.[28]), “Glossary: Child deprivation”, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Glossary:Child_deprivation.

Child material deprivation extends beyond income poverty

The relative risk of material deprivation is higher for children in income‑poor households. On average across European OECD countries, children living in income‑poor families are nearly five times more likely than their non-income‑poor peers to experience child-specific material deprivation. However, most children who experience material deprivation do not come from income‑poor households. In 2024, of the 10.3% of children experiencing material deprivation, roughly 3.3% were income‑poor, while around 7% were not (Figure 3.7, Panel A). In other words, non-income‑poor children account for about two‑thirds of all materially deprived children in European OECD countries, with this proportion reaching 89% in Finland.

Figure 3.7, Panel B shows that material deprivation in many countries also extends beyond the lowest-income households. On average across OECD European countries, around 30% of materially deprived children live in households with incomes above the third decile of the income distribution. This share reaches around 50% in Lithuania and Latvia, while in Luxembourg and Czechia it lies at 12% of children.

Figure 3.7. Material deprivation extends beyond income poverty
Copy link to Figure 3.7. Material deprivation extends beyond income poverty

Child poverty is more prevalent in single‑parent and large families

Not all family types face the same risk of income poverty and material deprivation. Children living in single‑parent and large families are significantly more likely than others to experience these forms of disadvantage. Figure 3.8, Panel A shows that children living in single‑parent households face a markedly higher risk of income poverty and child-specific material deprivation than with two adults (see Annex 3.A for data by country). Exposure to poverty and material deprivation also increases with the number of children in the household. On average across the OECD, children living in families with three or more children are twice as likely to experience income poverty and material deprivation as those living in families with only one child. These patterns are consistent with evidence showing that single‑parent and large families face disproportionately higher costs of raising children, whether measured through household consumption expenditures or self-reported financial strain (Rapp and Thévenon, 2025[29]; OECD, 2025[30]).

Beyond differences by family structure, Figure 3.8, Panel B also documents age‑related patterns. In line with the standard European indicator of child material deprivation – defined as an enforced lack of at least three out of 17 essential items – 13.6% of children under age 16 in the European Union were materially deprived in 2024. In several countries, young adolescents aged 12‑15 are more frequently exposed to deprivation than younger children, notably in Austria (+7.4 p.p. relative to pre‑school children), Ireland (+6.0 points), and Lithuania (+4.9 points). By contrast, in Norway, Finland, the Netherlands, the Slovak Republic, Portugal and Hungary, material deprivation is lowest among children aged 12‑15, diverging from the prevailing pattern.

The age gradient of child poverty reflects how family circumstances and policy support evolve over the life course. Younger children can be at greater risk of poverty resulting from low household market income because their parents tend to have lower and less stable earnings early in their careers and may reduce work due to childcare responsibilities. As children grow older, parental earnings typically increase and family situations stabilise, which can reduce poverty risks. At the same time, the design of public support matters: differences in how benefits and services are targeted across age groups, combined with changing costs of children (e.g. childcare vs. education), can shift poverty risks between younger and older children. Figure 3.8, Panel B suggests that in most countries, social policies protect young children from material deprivation relatively well.

Figure 3.8. Children in single‑parent and large families are at higher risk of income poverty and material deprivation
Copy link to Figure 3.8. Children in single‑parent and large families are at higher risk of income poverty and material deprivation

3.2. Child poverty reduction strategies as a social investment approach

Copy link to 3.2. Child poverty reduction strategies as a social investment approach

As discussed in Chapter 2, several aspects of the evolution of social policies point to a shift – whether explicit or not always formally articulated – towards the principles underlying a social investment approach. Policies aimed at tackling child and family poverty are among those increasingly guided by the objective of providing children (and their parents) with opportunities to realise their full potential, while supporting children’s development, well-being, and the accumulation of human and social capital – central goals of the social investment perspective.

To this end, child poverty policies can operate through both preventive and mitigating channels. Broadly, preventive measures aim to reduce families’ reliance on income transfers and public services by supporting parents’ access to employment and strengthening their earning potential. Mitigating measures, by contrast, provide cash and in-kind transfers that cushion the effects of low income and material deprivation, helping families remain above the poverty threshold or reducing the depth of poverty among those who remain poor. In practice, the distinction between preventative and mitigating policies is not always clear cut. For example, early childhood education and care can enable parents to increase their employment whilst also supporting child development, particularly in low-income families.

Many countries in the OECD and beyond have shown strong commitments to reducing child poverty but their approaches differ in terms of the policy levers they prioritise and the comprehensiveness of policy strategies (Global coalition to end child poverty, 2025[32]; OECD, 2018[33]). A notable example of a longstanding and comprehensive commitment is Canada, where successive governments have implemented policies and programmes to support families with children, often incorporating progressive adjustments to increase financial assistance for low- and modest-income households (Thévenon, 2018[34]). These efforts were further strengthened with the launch in 2018 of the Opportunity for All – Canada’s First Poverty Reduction Strategy, which called for a co‑ordinated approach with diverse groups, government and non-government alike (Government of Canada, 2018[35]). Key measures included strengthening income support programmes, expanding employment and training opportunities, developing affordable housing, enhancing the education and skills of disadvantaged children, improving access to healthcare and social services, and bolstering community-based initiatives.

New Zealand offers another example of sustained action to combat child poverty. The Child Poverty Reduction Act of 2018 established mandatory benchmarks requiring the government to set and report on three‑year and ten‑year targets across multiple indicators, including income poverty (before and after housing costs) and material hardship. Building on this foundation, the Child and Youth Strategy 2024‑2027 (Government of New Zealand, 2024[36]) set out three priorities for children and families: i) supporting children and their families and whānau in the first 2 000 days; ii) reducing child material hardship by increasing household incomes, supporting access to stable housing, lowering essential living costs and addressing long-term disadvantage; and iii) preventing harm against children. A distinctive feature of New Zealand’s approach is its integration into a social investment approach, under which policies to break cycles of disadvantage are guided by evidence on community needs and effective interventions. This approach emphasises empowering communities and ensuring that resources are spent efficiently, with a clear focus on delivering measurable value for money (Government of New Zealand, 2025[37]).

Child poverty has been a recurring focus of UK policy over the past decade. The Child Poverty Act 2010 initially established legally binding income‑based targets, but these were largely dismantled after 2015, leaving child poverty monitored through statistics rather than enforceable commitments. As child poverty rose to record levels – reaching almost 16% of children in 2023 (Chapter 2) – criticism grew that incremental welfare adjustments were insufficient. In December 2025, the UK Government introduced a new national Child Poverty Strategy, signalling a shift toward a more co‑ordinated and structural approach (HM Government, 2025[38]). Central to the strategy is the abolition of the “two-child limit” on Universal Credit (income support, housing benefits and tax credits), a reform expected to lift hundreds of thousands of children out of relative poverty. The strategy also expands free school meals, extends early years childcare for low-income families, subsidises essential goods such as infant formula, addresses temporary accommodation for homeless families, and strengthens local support through Best Start Family Hubs. Overall, the 2025 reframing moves from piecemeal measures toward a broader package combining income support with cost reduction and service access, though its long-term impact will depend on sustained funding and political commitment.

In the European Union, the question of whether child poverty reduction policies can be effectively guided by a social investment approach has been part of the debate since the early 2000s, focussing on how countries can reform welfare systems and align EU member states around shared priorities to enhance the impact of social spending (Esping-Andersen, Gosta, 2002[39]; Vandenbroucke, Hemerijck and Palier, 2011[40]; Bouget et al., 2015[41]; European Commission, 2024[42]). Several EU-level initiatives adopted since the late 1990s have been instrumental in reducing the risk that children grow up in poverty or suffer from its consequences. This includes the European Employment Strategy, which promotes measures to help parents reconcile work and family life and supports maternal employment, notably through guarantees of paid leave and targets for the provision of early childhood care and services (Thévenon, 2004[43]). Further action was reinforced through a series of EU initiatives explicitly aimed at addressing child poverty and investing in children, with the overarching goal of breaking the intergenerational transmission of disadvantage (Box 3.2).

Box 3.2. European Union initiatives to address child poverty and invest in children

Copy link to Box 3.2. European Union initiatives to address child poverty and invest in children

Since the early 2010s, several EU-level initiatives have sought to encourage Member States to take action against child poverty and to ensure that children living in families at risk of poverty or social exclusion have access to quality services that can help break the cycle of disadvantage.

In 2013, as part of the Social Investment Package (n.d.[44]), the European Commission issued the Recommendation on Investing in Children: Breaking the Cycle of Disadvantage (2013[45]). This Recommendation calls on Member States to prioritise children’s rights and well-being by ensuring access to adequate resources, affordable quality services, and opportunities for child participation. It emphasises integrated strategies that combine income support for families, inclusive labour markets, and high-quality education, health, housing, and childcare services. The Recommendation also highlights the importance of early intervention and prevention, focussing on tackling inequality and supporting disadvantaged children to break intergenerational cycles of poverty and social exclusion. In 2017, the Commission, together with experts from the European Social Policy Network, prepared an evaluation of the Recommendation’s implementation.

Several principles of the European Pillar of Social Rights (n.d.[46]), proclaimed in 2017, support breaking the cycle of disadvantage. These include the right to quality and inclusive education and lifelong learning (Principle 1); fostering equal opportunities for under-represented groups (Principle 3); and, for children, access to affordable early childhood education and care, protection from poverty, and measures to improve opportunities for disadvantaged children (Principle 11). In addition, timely access to quality healthcare (Principle 16) and access to social housing or housing assistance for those in need (Principle 19) are recognised as essential for promoting equality and social inclusion.

The European Pillar of Social Rights Action Plan (n.d.[47]), adopted in 2021, sets a target to reduce the number of people at risk of poverty or social exclusion by at least 15 million by 2030, of which at least 5 million should be children. Focusing on children not only provides them with access to new opportunities but also contributes to breaking the intergenerational cycle of poverty, preventing them from becoming adults at risk of poverty or social exclusion, and producing long-term systemic effects.

Building on the above principles and as announced in the European Pillar of Social Rights Action Plan (n.d.[47]) and in the EU Strategy on the Rights of the Child, the Council established the European Child Guarantee (ECG) in 2021 (2021[48]). The ECG is part of a broader set of integrated measures outlined in the Action Plan and aims to ensure that children at risk of poverty or social exclusion have access to fundamental rights, including healthcare and education. The ECG recommends that Member States guarantee children in need effective and free access to early childhood education and care, school-based education and activities, at least one healthy meal every school day, healthcare, healthy nutrition, and adequate housing. Beyond access, the ECG emphasises high-quality and inclusive services, focussing on availability, affordability, timeliness, and public awareness of children’s rights to use these services.

Under the ECG, each Member State has appointed a national Child Guarantee Co‑ordinator to oversee and monitor implementation, including assessing children in need and the barriers they face in accessing services. By 2030, Member States are asked to develop and implement national action plans, grounded in an integrated policy framework to combat child social exclusion. These plans should align with broader social, education, health, nutrition, and housing policies; increase investment in children; provide support measures for families; address geographical disparities; strengthen co‑operation with stakeholders (including children); allocate sufficient resources and EU funds; and integrate a gender perspective.

In May 2026, the European Commission has unveiled a major social policy package aimed at reducing poverty, tackling housing exclusion, strengthening children’s opportunities, and enhancing the rights of persons with disabilities across the EU. The package includes the EU’s first Anti-Poverty Strategy (n.d.[49]), a proposal for a Council Recommendation on fighting housing exclusion (2026[50]), a Communication on breaking the cycle of child poverty (2026[51]) – strengthening the European Child Guarantee (n.d.[52]), and a Communication on enhancing the strategy for the rights of persons with disabilities up to 2030 (n.d.[53]). The Anti-Poverty Strategy seeks to reduce the number of people at risk of poverty or social exclusion by at least 15 million by 2030 and ultimately to eradicate poverty by 2050. Key priorities include promoting quality employment, improving access to services and adequate income support, and strengthening co‑ordination across policies. The package also places strong emphasis on children, recognising that one in four children in the EU remains at risk of poverty or social exclusion and underscoring the importance of early-life interventions to break intergenerational disadvantage. Strengthening the European Child Guarantee is motivated by children’s persistently higher risk of poverty or social exclusion compared to adults, with past progress deemed insufficient. The aim is to embed policy efforts to reduce child poverty within a broader policy framework, underpinned by comprehensive approaches across policy areas (e.g. education, healthcare, housing) and by targeted measures tailored to the specific needs of vulnerable groups. To this end, the Communication outlines three pillars: securing parents’ access to quality jobs, childcare, and strong safety nets; ensuring vulnerable children’s access to essential services; and increasing and better co‑ordinating investment in children.

Source: European Commission (2026[54]), “Commission proposes new measures to fight poverty and improve lives of persons with disabilities”, https://employment-social-affairs.ec.europa.eu/news/commission-proposes-new-measures-fight-poverty-and-improve-lives-persons-disabilities-2026-05-06_en.

The extent to which recent social policy developments have contributed to reducing child poverty – or preventing larger increases – while also helping to contain growth in social spending, remains an open question. Understanding to what extent, under what conditions, and through which policy designs social expenditures are most effective in reducing child poverty while keeping spending pressures under control is therefore essential for informing future policy strategies. These issues are examined in the following sections.

3.3. The methodology

Copy link to 3.3. The methodology

The analysis goes beyond assessing whether higher social spending is associated with lower child poverty. It examines whether the design and allocation of social spending can influence the value for money, by identifying policy configurations associated with stronger poverty reduction and lower long-term spending pressures. It builds on earlier work by Lorenzoni et al. (2018[55]) that examined the impact of healthcare policies and institutional arrangements on life expectancy and health spending, and by Égert et al. (2020[56]) that investigated how education policies affect public education spending and the accumulation of human capital.

In this framework, value for money is assessed along two complementary dimensions. The first captures the effectiveness of social spending in improving child outcomes, measured here by reductions in child poverty and severe material deprivation. The second captures long-term spending pressure, namely whether particular policy designs are associated with stronger or weaker growth in social expenditure. The analysis therefore seeks to identify policy choices that are associated with a stronger poverty-reducing effect of social spending while also limiting longer-term pressures on public spending.

The analysis is deliberately narrowed to child poverty outcomes to determine the value for money of social policy design alongside the impact on the long-term evolution of social spending. Social policies generate many other benefits for children – including improvements in health and learning outcomes – and pursue wider social objectives that are not captured by this empirical framework. Some of these broader effects are examined separately in the following chapter.

The empirical approach estimates the two value for money dimensions separately but links them through social spending. The first equation estimates the influence of social policy design on the long-run relationship between per capita social expenditure and child poverty, while accounting for demographic and structural factors that also shape poverty risks. Importantly, it allows the influence of policy design on the poverty-reducing effect of social spending to vary according to the level of social expenditure. This makes it possible to assess whether particular policy characteristics are associated with stronger or weaker effects depending on the size of the welfare state.

Child poverty is measured by three complementary variables:

Child relative poverty rate, measuring the share of children considered at risk of poverty, i.e. those living in households with equivalised disposable income that is below 50% of the national median. The threshold rises with annual increases in the median income.
Child anchored poverty rate, capturing the share of children who would be considered at risk of poverty based on living standards in the past, namely those below the inflation-adjusted poverty threshold “anchored” in 2005. As living standards have increased over time in most OECD countries, the anchored poverty rate allows to focus on a subset of children who live in deeper levels of poverty compared to all those falling under the relative poverty line.
Severe material deprivation rate among children, measuring the share of children living in households that experience an enforced inability to afford some items considered by most people to be desirable or even necessary to lead an adequate life. This indicator covers nine core aspects of material well-being, including the ability to keep the home adequately warm, afford nutritionally adequate meals, cope with unexpected expenses, and afford a one‑week annual holiday away from home4 (see Annex 3.B for a detailed definition).

The second equation estimates the long-run determinants of per capita social spending, allowing the relationship between countries’ economic and social spending growth to vary with social policy design features and controlling for demographic factors. There are several channels through which policy design can affect the longer-term evolution of social expenditure. Some policy features may strengthen families’ labour market attachment and economic self-sufficiency, potentially reducing reliance on income support and moderating future spending pressures. Other policy designs may be less substitutable with other spending areas and thus associated with stronger growth in social expenditure over time. Taken together, the two equations provide a basis for assessing both the effectiveness of social spending in reducing child poverty and the spending implications under alternative policy configurations. The technical details are provided in Box 3.3.

The policy variables capture key design characteristics of social expenditure systems and several dimensions of policy choice, including the allocation of spending across policy areas; the balance between cash transfers and services; employment-oriented policies; and the targeting of income support towards lower-income households. The analysis thus considers, for example, whether allocating a greater share of social expenditure to family policies or employment-oriented policies is associated with stronger poverty reduction; whether support delivered through services, such as ECEC and healthcare, is associated with different outcomes from cash support; and whether greater targeting of income transfers towards lower-income households is associated with stronger poverty-reducing effects.

The policy variables cover both policies directed specifically towards families and children and broader social protection programmes. This reflects the fact that many programmes that are not targeted specifically at parents or children nevertheless affect children’s living standards, while accounting for a large share of total social expenditure. Since the policy characteristics are measured in different units, they are standardised before estimation to allow their estimated associations to be compared on a common scale. A coefficient on a policy variable consequently indicates the estimated association with child poverty or social spending of a one‑standard-deviation difference in the relevant policy characteristic, based on the historical variation observed across countries and over time. This allows, for example, the estimated association of greater targeting of income transfers to be compared with that of higher participation in ECEC, despite the two indicators being measured on different scales. The estimates should be interpreted as comparisons of the relative importance of policy characteristics of similar magnitude within their observed distributions, rather than as estimates of the effect of an identical investment or p.p. change in each policy variable.

The specific policy indicators included in the analysis are outlined below, and further details on variable definitions and data sources are provided in Annex 3.B.

Box 3.3. How is the long-run value for money of social policies estimated?

Copy link to Box 3.3. How is the long-run value for money of social policies estimated?

The empirical framework consists of two long-run equations estimated independently using dynamic ordinary least squares (DOLS) that assess the two dimensions of value for money:

1. Effectiveness: The first equation relates child poverty outcomes to per capita social expenditure and policy design to assess whether a social policy characteristic strengthens or weakens the estimated relationship between social expenditure and child poverty or severe material deprivation.

2. Spending moderation/pressure: The second equation relates social expenditure to economic development and policy design to identify whether a social policy characteristic is associated with a stronger or weaker relationship between structural economic pressures and the level of social expenditure.

The framework is designed to identify policy configurations associated with greater poverty-reducing effectiveness of social spending and/or lower long-run spending pressure over time, rather than to estimate the fiscal cost of a specific social policy or the poverty-reducing effect of a budget-neutral reallocation of social expenditure in the short run.

Equation 1: Child poverty and the effectiveness of social spending

For country i and year t, the child outcome is specified as:

${C O}_{i t} = α_{i} + μ_{t} + {T S}_{i t} β + P_{i t}^{T} γ + {T S}_{i t} P_{i t}^{T} δ + {X O}_{i t}^{T} η + \sum_{k = - 1}^{1} Δ {L L O}_{i, t + k}^{T} ζ_{k} + ε_{i t} (1)$

where:

$C O_{i t}$ is the child outcome. Three outcomes are considered separately: the relative child poverty rate, the anchored child poverty rate, and the severe material deprivation rate among children;
${T S}_{i t}$ is total social expenditure per capita, expressed in constant USD PPP and in logarithmic form;
$P_{i t}$ is a vector of social policy design characteristics;
${T S}_{i t} P_{i t}^{T}$ captures whether the influence of policy design features on the effectiveness of social spending in reducing child poverty varies with the size of the welfare state;
${X O}_{i t}$ is a vector of demographic and core drivers of child poverty;
$α_{i}$ and $μ_{t}$ denote country and time effects.

Total social expenditure includes public and private expenditure in eight social policy areas: family, health, incapacity-related benefits, unemployment, active labour market programmes, housing, survivors and other social policy areas. Spending on old-age pensions is excluded because it is not directly targeted at families or children and its effects on children’s living standards are ambiguous.1 Its large weight in total social expenditure could also obscure the relationship between spending in child-relevant policy areas and child outcomes.

The vector $P_{i t}$ includes indicators capturing the allocation and design of social policy, including the composition of social expenditure across policy areas, participation in ECEC, employment-related policy characteristics, the targeting of income support, and other relevant policy features examined in the analysis. The expenditure share on active labour market policies (ALMP) is lagged by three years to reflect the time needed for employment-oriented interventions to affect household economic circumstances (see (Card, Kluve and Weber, 2017[57])). A full description of the policy variables can be found in Annex 3.B.

The core controls in ${X O}_{i t}$ capture changes in demographic structure and the depth of market-income poverty.2 They include the average number of children per household,3 the adolescent fertility rate,4 the share of single‑parent families5 and the mean poverty gap before taxes and transfers.6

Equation 2: Drivers of social spending and spending pressure

The second equation models long-run per capita social expenditure growth, assuming that exogenous, structural macroeconomic and demographic drivers determine the level of per capita social spending by shaping a country’s capacity and constraints to spend, as:

${T S}_{i t} = ρ_{i} + σ_{t} + {G D P}_{i t} θ + P_{i t}^{T} κ + {G D P}_{i t} P_{i t}^{T} λ + {X S}_{i t}^{T} φ + \sum_{k = - 1}^{1} Δ {L L S}_{i, t + k}^{T} ξ_{k} + υ_{i t} (2)$

where:

${G D P}_{i t}$ is GDP per capita net of social expenditure per capita, expressed in constant USD PPP and in logarithmic form. It represents countries’ financial resources and capacity for social spending and is hence positively associated with the level of social expenditure per capita;
${X S}_{i t}$ captures other core drivers of social spending, notably life expectancy7 and the dependency ratio,8
$P_{i t}$ represents the same set of social policy design characteristics as in Equation 1;
the interaction between policy characteristics and economic resources ( ${G D P}_{i t} P_{i t}^{T}$ ) allows the influence of social policy characteristics to vary according to a country’s level of economic development;
$ρ_{i}$ and $σ_{t}$ denote country and time effects.

The equation therefore separates the role of core factors that shape countries’ capacity and pressures to spend from the role of policy choices that can amplify or moderate these spending relationships.

Estimation and interpretation

The modelling assumes that, in the long run, structural macroeconomic and demographic drivers shape countries’ capacity and constraints to spend, and, in turn determine the level of per capita social spending. Social policy choices (denoted by the vector $P_{i t}$ ) amplify or attenuate the long-run, positive relationship between GDP and social expenditure (equation 2). Social expenditure, in turn, affects the child poverty outcome ( $C O_{i t}$ ), alongside other core demographic and economic drivers (vector ${X O}_{i t}$ ), as captured by equation (1). Importantly, policy design may also influence the effectiveness of social spending in reducing child poverty, such that the same level of social expenditure can be associated with different child poverty outcomes depending on how social policies are designed (equation 1).

Modelling the determinants of child poverty, equation (1) separates the average effect of increasing the overall per capita social budget ( $β$ ), from the returns to expanding specific policies ( $γ$ ) as well as potential variations in policies’ effectiveness for different levels of welfare state generosity ( $δ$ ). Analogously, equation (2) allows to distinguish between the impact of policies on spending moderation/pressures according to different stages of economic development.

Both equations are estimated separately using dynamic ordinary least squares (DOLS). DOLS augments the long-run relationship with leads and lags of first differences of the explanatory variables ( $Δ {L L O}_{i, t + k}$ and $Δ {L L S}_{i, t + k}$ , respectively). This helps account for short-run dynamics, including contemporaneous endogeneity, serial correlation and adjustment processes that may otherwise bias estimates of long-run relationships. The limited length of the available time series restricts the number of leads and lags that can be included to three time periods ( $t - 1$ , $t$ and $t + 1$ ).

Although not explicitly modelled, the framework is consistent with the assumption that over time, improvements in child outcomes will translate into enhanced human capital and productivity when they are adults, which results in continued GDP growth.9

Value‑for-money simulations

The value‑for-money simulations use the long-run coefficients estimated in Equations (1) and (2) to simulate the incremental effect of changing a specific policy characteristic when overall social expenditure or GDP per capita also increases. The simulations therefore assess the contribution of policy design over and above the effect of higher overall spending or economic growth.

Effectiveness in improving child outcomes

For a given policy characteristic $p_{j}$ Equation (1) can be used to calculate the change in the child outcome associated with increasing the policy alongside an increase in per capita social expenditure. The simulated outcome change is:

${O u t c o m e c h a n g e}_{j} = \hat{γ_{j}} * Δ p_{j} + \hat{δ_{j}} * Δ p_{j} * (T S + Δ T S) (3)$

where $Δ p_{j}$ represents a simulated increase in the policy characteristic and $Δ T S$ a simulated increase in per capita total social expenditure. In the simulations, $Δ T S$ corresponds to an increase of half a standard deviation in the log of per capita total social expenditure. Equation (3) therefore captures the additional change in the child outcome associated with changing the policy characteristic together with increasing social expenditures, relative to a counterfactual scenario in which social expenditure increases by the same amount but the policy characteristic remains unchanged.10 The relevant coefficients are taken from the regression specification detailed in Equation (1) where the values for $\hat{γ_{j}}$ and $\hat{δ_{j}}$ are taken from the regression estimates related to policy $j$ . Statistically insignificant coefficients are set to zero when calculating the simulated effects.

A negative ${O u t c o m e c h a n g e}_{j}$ indicates that the policy change is associated with an additional reduction in the child poverty or material deprivation rate relative to the counterfactual. A positive value indicates that the policy change reduces the estimated poverty-reducing effect of the increase in social expenditure. It does not necessarily mean that the policy increases the absolute level of child poverty, since the overall relationship between higher social expenditure and child poverty may remain negative.

Spending pressure

The corresponding simulation based on Equation (2) measures the additional spending pressure associated with changing a policy characteristic for a given level of GDP per capita and an increase by half a standard deviation ( $G D P + Δ G D P$ ):

${S p e n d i n g p r e s s u r e}_{j} = \hat{κ_{j}} * Δ p_{j} + \hat{λ_{j}} * Δ p_{j} * (G D P + Δ G D P) (4)$

Based on the estimated regression coefficients $\hat{κ_{j}}$ and $\hat{λ_{j}}$ related to policy $p_{j}$ from equation (2), the simulation compares a scenario in which GDP per capita and the policy characteristic both increase with a counterfactual scenario in which GDP per capita increases by the same amount but the policy characteristic remains unchanged.11

A negative value indicates spending moderation, meaning that the projected increase in social expenditure is smaller than in the counterfactual. A positive value indicates additional spending pressure. A negative value should not be interpreted as a reduction in the absolute level of social expenditure.

Magnitude of policy simulations

Two types of policy changes are considered.

Standardised policy changes. For the main set of simulations that spans all policies, the policy characteristic is increased by one standard deviation, while log per capita social expenditure or GDP per capita increases by half a standard deviation. The one‑standard-deviation policy change reflects the historical variation of each policy characteristic across countries and over time and provides a common metric for policy variables measured in different units. The half-standard-deviation increase corresponds to approximately 4% of the sample average for log per capita social expenditure and approximately 2% for log GDP per capita, broadly equivalent to around 11 years of average growth in the sample. These simulations are therefore intended to represent policy changes within the range of historical OECD experience rather than marginal changes of an identical absolute size.
Spending-allocation changes. For policy variables defined as the share of total social expenditure allocated to a specific policy area, a second set of simulations is performed. Here, the policy increase represents one p.p. of social expenditure. Because the change is identical across policy areas, this simulation provides a more direct comparison of the estimated implications of allocating additional resources to different areas of social spending. It is particularly relevant for considering how additional resources might be allocated across competing policy areas.

Interpreting value for money

Based on equations (3) and (4), the “value for money” of policy $j$ is then given jointly by its child outcome change and spending pressure. This approach captures the efficiency of policy design rather than absolute policy effects. The objective is not simply to assess whether a policy characteristic influences outcomes, but whether it improves the overall effectiveness of social spending when expenditure is assumed to grow over time. In other words, it examines whether changing the composition or design of social expenditure generates better child outcomes than increasing spending alone. Conceptually, the approach addresses the following question: assuming that governments increase per capita social spending as economies grow, which policy designs are associated with larger improvements in child outcomes without proportionally increasing long-run spending pressures?

The two simulated effects are considered jointly:

	*Spending pressure moderated or unchanged*	*Spending pressure increased*
*Additional reduction in child poverty*	Relatively strong value for money	Trade‑off: stronger child outcomes but greater spending pressure
*No additional reduction/ weaker poverty reduction*	Trade‑off: lower spending pressure but weaker child-outcome effects	Relatively weak value for money

The classification is based on the relative performance of a policy change against the relevant counterfactual of no policy change, rather than on whether the policy has a positive or negative effect in absolute terms.

Where the interaction between policy design and the core driver is statistically significant, the simulated value‑for-money effect varies with the level of social expenditure or GDP per capita. The same policy characteristic may therefore have different estimated implications in countries at different levels of economic development or with different levels of social spending.

The simulations should not be interpreted as forecasts or as estimates of the effects of a specific budget-neutral reform. In particular, a one‑p.p. increase in the share of spending allocated to one policy area is not assumed to be financed by an equivalent reduction in another area. The simulations instead isolate the estimated role of policy design conditional on an overall increase in social expenditure or economic resources.

It should be noted that the “value for money” estimates offer important insights to increase the efficacy of social expenditure in improving child outcomes, but they do not capture the total value of social policies. First, the analysis focusses on a limited set of child outcomes and is therefore likely to overlook additional, unmeasured benefits accruing during childhood. For example, increased employment and demanding work schedules may affect the quality of parental care, ECEC can support children’s well-being and development in addition to reducing the share of children living at risk of income poverty, and public healthcare systems not only benefit parental employment but also the physical and mental well-being of children themselves. Second, a key objective of social policies targeting children is not only to improve well-being during childhood itself, but also to strengthen outcomes later in life. In particular, improvements in childhood conditions – including reductions in poverty – have been shown to yield benefits later in life, such as better employment prospects, higher earnings and improved health outcomes (Clarke et al., 2022[3]). These lifelong impacts cannot be fully captured because of the limited length of the available time series. As a result, the present estimates are likely to underestimate the full long-term benefits of investments in children.

1. Retirement pensions and other spending on old age are excluded for several reasons. First, they are not a policy instrument that governments intentionally use to address child poverty; while they may have indirect effects, these are ambiguous, and the available do not allow us to capture the underlying mechanisms. Second, pension spending is so large that it risks dominating overall social expenditure trends, making it harder to control for unobserved factors, which may in turn bias estimates of the impact of social spending on child poverty. If old age benefits are mainly received by households without children, they can raise the poverty threshold more than children’s household incomes, thereby increasing measured child poverty. Conversely, when pension benefits accrue to multigenerational households where pension income is pooled with other household resources and may therefore benefit children, they can lift families above the poverty line. Empirical evidence suggests the latter effect dominates in countries where extended family networks play a central role in welfare provision – particularly in Southern and Central and Eastern Europe – where pensions significantly reduce child poverty (Petmesidou et al., 2016[58]; Diris, Vandenbroucke and Verbist, 2017[59]). In these contexts, pension cuts can indirectly increase child poverty. By contrast, in countries where pension spending is largely directed to households without children, higher pension outlays may be associated with higher child poverty.

2. GDP per capita is not included as a separate explanatory variable in the child outcomes equation to avoid multicollinearity with other drivers, especially total social expenditure, and because doing so would make it difficult to disentangle its effects from those operating through social spending and other drivers within this empirical framework.

3. Average number of children per household. A decline in fertility reduces the average number of children per household, which can lower the overall risk of child poverty when poverty is concentrated in larger families. However, this effect is weaker if larger families remain most common among households with fewer economic resources, while fertility declines primarily among more advantaged groups (Thévenon et al., 2018[8]). The postponement of family formation associated with lower fertility may further reduce poverty risks, as couples often delay having children until achieving greater labour market stability and income security. This protective effect also depends critically on whether lower socio‑economic groups postpone parenthood to the same extent as higher-income groups.

4. Adolescent fertility rate. A clearer positive trend is the decline in adolescent fertility, which is strongly associated with higher educational attainment and greater labour market attachment (Schulkind and Sandler, 2019[60]; Hobcraft and Kiernan, 2001[61]), and therefore likely to contribute to reduced child poverty.

5. Share of single‑parent families. Family dissolution often reduces household resources and single parents continue to bear a disproportionately high cost of children compared to two‑parent households, – whether measured through consumption patterns or self-reported difficulties in making ends meet – making single‑parent households disproportionately vulnerable to poverty (Rapp and Thévenon, 2025[29]). The rising share of such families has consequently been an important structural driver of child poverty in many OECD countries (Thévenon et al., 2018[8]; Maldonado and Nieuwenhuis, 2015[62]).

6. Mean poverty gap before taxes and transfers. The percentage gap between the equivalised income before taxes and benefits of the average individual living in a working-age household and the relative poverty line is a measure of market inequalities, the average “depth” of poverty before public transfers. Deeper forms of poverty before taxes and benefits and can diminish the child poverty-reducing effect of social policies, for example when cash transfers are insufficient in size to cover the income gap relative to the poverty line.

7. Life expectancy enters in logged terms as part of the other core drivers, ${X S}_{i t}$ . Increases in life expectancy have had a positive impact on social spending in OECD countries over the past decades. Greater longevity has not fully translated into gains in healthy life expectancy and was often achieved by advances in medicine and incurred expensive treatment costs (OECD, 2025[63]). In addition, where greater life expectancy translates into longer periods spent in the labour market, it can bring additional tax revenues and make it more affordable for countries to increase social expenditure.

8. Dependency ratio: Another core driver captured by ${X S}_{i t}$ , the ratio of children and the elderly – those below 20 and those 65 years or older – to the working-age population influences per capita social spending by limiting the ability to finance per person social expenditures.

9. The implicit assumption that lower child poverty spurs economic growth relies mainly on improved outcomes in adulthood, such as higher employment, better health, and greater productivity (Clarke et al., 2022[3]), which are key channels through which social investment policies are expected to generate long-term returns.

10. Note that equation (3) is obtained by taking the difference between the predicted outcome under a scenario in which the increase in total social expenditure is accompanied by a change in the relevant policy characteristic and the counterfactual scenario in which total social expenditure increases but the policy mix remains unchanged. Formally: ${O u t c o m e c h a n g e}_{j} = \hat{β} * (T S + Δ T S) + \hat{γ_{j}} * (p_{j} + Δ p_{j}) + \hat{δ_{j}} * (p_{j} + Δ p_{j}) * (T S + Δ T S) - (\hat{β} * (T S + Δ T S) + \hat{γ_{j}} * (p_{j}) + \hat{δ_{j}} * (p_{j}) * (T S + Δ T S))$

11. Note that equation (4) is equivalent to: ${S p e n d i n g p r e s s u r e}_{j} = \hat{θ_{j}} * (G D P + Δ G D P) + \hat{κ_{j}} * (p_{j} + Δ p_{j}) + \hat{λ_{j}} * (p_{j} + Δ p_{j}) * (G D P + Δ G D P) - (\hat{θ_{j}} * (G D P + Δ G D P) + \hat{κ_{j}} * (p_{j}) + \hat{λ_{j}} * (p_{j}) * (G D P + Δ G D P))$

Social policy leavers to address child poverty and material deprivation

Policy features ( $P_{i t}$ ) capture key design characteristics of social expenditure systems that are assumed to influence both the effectiveness of social spending in reducing child poverty and the way social expenditures evolve in relation to GDP per capita. The set of policy levers allows to explore key choices policymakers face when allocating public resources, including whether to prioritise spending within the family policy domain or through employment-oriented policies; whether support is more effective when delivered in cash or through services (such as childcare, work-life reconciliation measures, or health and social services); whether cash transfers are more impactful when delivered through benefits or tax breaks; and the extent to which targeting support toward low-income households enhances poverty-reduction outcomes. The analysed policy areas comprise elements of social protection and interventions financed through social expenditure that do not target parents or children specifically, as these programmes account for the bulk of total social spending and can play an important role in reducing child poverty. Some policy variables indicate the share of social expenditure dedicated to a specific policy area, irrespective of the size and type of the target population, to assess the average impact of budget allocations on child poverty.5 Moreover, some of these policy features may – intentionally or not – strengthen families’ labour market attachment and self-sufficiency, thereby reducing reliance on benefits and moderating future spending pressures. Others on the other hand, may be drivers of social expenditure growth, resulting in higher long-term pressure on social expenditures.

The specific indicators included in the analysis are outlined below, and further details on variable definitions and data sources are provided in Annex 3.B.

Employment-related policies

Social policies are critical for supporting parents’ participation in paid work, increasing work intensity and boosting earnings, and thereby reducing child poverty (OECD, 2018[33]; Bäckman and Ferrarini, 2010[64]; Whiteford and Adema, 2007[65]). Investment in ECEC services is a particularly important lever for addressing child poverty in this respect, even if it is not the primary objective of these policies. On the one hand, affordable, high-quality childcare supports parents in meeting their care responsibilities whilst increasing their labour market participation and long-term earnings potential, thereby reducing poverty risks; on the other, it promotes children’s early development and learning (Adema, Clarke and Thévenon, 2020[66]; Hemerijck, Ronchi and Plavgo, 2023[67]). In addition, employment-oriented policies combine a wide range of services that help parents (as well as non-parents) enter, remain in, or progress in employment with financial incentives to take up work. Guidance and mentoring programmes, including support to identify and access childcare arrangements that make employment compatible with family life, as well as training and upskilling programmes that facilitate transitions into better-quality jobs, are especially important. More intensive and tailored services targeted at hard-to-place workers are particularly relevant for low-skilled parents facing multiple barriers to employment. Further, low effective tax rates on earned income can encourage parental labour supply and boost disposable household income by limiting the share of additional earnings repaid as taxes and social contributions. The set of indicators that describe policies with the potential to encourage and facilitate employment includes:

ECEC enrolment 0‑2: The share of children aged 0‑2 enrolled in formal care;
ECEC enrolment 3‑5: The share of children aged 3‑5 enrolled in early childhood education and care (ECEC) services;
Spending on ECEC: The share of per capita social expenditure allocated to ECEC services.
Spending on ALMP: The share of per capita social expenditure allocated to active labour market programmes (ALMP). They cover public spending on services and programmes aimed at improving employability and job matching – such as public employment services, job-search assistance, training and upskilling, employment incentives, supported employment and rehabilitation, and direct job creation – excluding passive income support for the unemployed. The variable is lagged by three years to account for the time it may take for these programmes to translated into greater employment (Card, Kluve and Weber, 2017[57]).
Effective tax rates on taking up work: Two indicators measure the financial disincentives to participate in the labour market for hypothetical families with two children aged 4 and 6. They calculate the proportion of earnings in the new job that are lost to either higher taxes or lower benefit entitlements when i) a jobless single parent takes up full-time employment at 67% of the average wage after claiming unemployment benefits and ii) one parent in a couple takes up full-time employment at 50% of the average wage after claiming unemployment benefits, while the other parent is always employed full-time at 50% of the average wage.
Effective tax rates on increasing working hours: Two indicators measure the fraction of additional earnings that is lost to either higher taxes or lower benefits for hypothetical families with two children aged 4 and 6 when i) a single parent increases their working hours from 50% to 100% of full-time work at earnings of 67% of the average wage and ii) one parent in a couple increases their working hours from 50% to 100% of full-time work at earnings of 50% of the average wage, while the other parent is always employed full-time at 50% of the average wage.

Other social policy expenditures in kind

Health-related social protection and services, as well as social services outside the family policy domain, can play a key role in addressing health problems, social difficulties, or other constraints that limit parents’ capacity to work. By helping the most disadvantaged parents to overcome these barriers and sustain employment, such interventions can have substantial downstream effects on children’s living standards and their risk of experiencing poverty. The corresponding policy features include.

Redistributive features of income transfers for the total population and for families specifically

Income transfers, through both cash benefits and tax6 relief, play a crucial role in protecting household incomes against economic shocks and in boosting resources to meet basic needs. Beyond their contribution to poverty reduction, income transfers serve different policy objectives and target different population groups. Some are designed to support households broadly, regardless of whether they have children, while others are targeted specifically at families with children to promote horizontal equity by compensating for the lower living standards associated with child-rearing relative to childless households (Box 3.4).

The extent to which these transfers contribute to reducing child poverty varies markedly across countries. This reflects policy trade‑offs between strengthening work incentives and providing adequate protection against poverty, as well as differences in the overall volume of support and in the extent to which redistribution targets low-income households or reallocates resources between childless households and families with children (Verbist and Van Lancker, 2016[68]; Thévenon et al., 2009[69]). Moreover, spending on families (including child and family allowances, maternity and parental leave benefits) may vary across different stages of childhood, with implications for children’s risk of poverty at different ages. These variations make it essential to consider carefully whether – and how – income transfer systems can be designed or adapted to strengthen their effectiveness in reducing child poverty. In this analysis, the main policy variables related to income redistribution capture the overall volume and distribution of redistribution, the allocation of cash support to families and children, and the design and progressivity of taxes and transfers.7

Box 3.4. How cash transfers for families can help reduce child poverty

Copy link to Box 3.4. How cash transfers for families can help reduce child poverty

Cash transfers are among the most direct policy instruments to reduce child poverty. By providing income support to families, they mitigate income shortfalls, help smooth consumption, and promote investment in children’s well-being and development. Their impact depends on benefit adequacy, coverage, and how they interact with tax and employment policies. Main mechanisms include:

Increasing disposable income. Cash benefits directly raise families’ resources, enabling them to meet essential needs such as food, housing, clothing and healthcare, and lifting many households above poverty thresholds.
Reducing income volatility. Regular transfers smooth fluctuations in household income – particularly for families facing unstable employment or low wages – helping to prevent poverty spells.
Supporting child-related costs. Family and child allowances compensate for the additional costs of raising children, promoting fairness between households with and without dependents.
Promoting investment in children. Increased household income enables greater spending on children’s nutrition, education and health, improving their long-term development and future earning capacity. Earmarking supports for children can be effective at increasing the household allocation of resources to children and to improve their material and developmental outcomes, particularly when financial constraints are high (Page, 2024[16]).
Complementing employment policies. Well-designed transfers can support parental employment – by reducing financial stress, helping with childcare costs, or being conditional on labour market participation – thus reinforcing the income effect.

Assessing value for money

To assess the value for money of different policy designs, we use the estimated relationships from the two value for money dimensions (Equations (1) and (2) in Box 3.3) to simulate how changes in individual policy characteristics would influence the long-run relationships between social spending and child poverty outcomes, and between GDP and social expenditure growth, as inspired by Egert et al. (2020[56]). The simulations compare two scenarios. In the first, overall per capita social expenditure (or GDP per capita, when assessing spending pressure) increases while a specific policy characteristic also changes. In the second, the underlying spending or GDP increase occurs but the policy characteristic remains unchanged. The difference between the two scenarios captures the incremental contribution associated with the change in policy design, conditional on the overall increase in spending or economic resources.

The simulations provide two complementary measures to assess the value for money. The first measures the change in the child outcome associated with strengthening a given policy alongside an increase in overall social spending. A negative value indicates that the policy change is associated with an additional reduction in child poverty or severe material deprivation compared with increasing social spending alone. A positive value indicates that the policy change weakens the poverty-reducing effect of the additional social spending, although it does not necessarily imply an increase in child poverty.

The second measure captures the additional spending pressure associated with changing a policy as economies grow. It compares the projected change in social expenditure when both GDP per capita and the policy characteristic increase with the projected change when GDP increases but the policy characteristic remains unchanged. A negative value indicates spending moderation: social expenditure is projected to increase less than it would in the absence of the policy change. Importantly, spending moderation does not imply that absolute social expenditure falls; it means only that its projected increase is smaller due to the policy change.

Considering policy changes alongside an increase in overall social spending or GDP reflects the long-run perspective of the analysis. Social expenditure tends to increase as economies grow, and many policy changes cannot realistically be implemented without additional resources. For example, expanding ECEC enrolment is likely to require additional expenditure that may not necessarily be financed through reduced investment in other social policies. Similarly, when a policy variable represents the share of social spending allocated to a particular policy area, increasing the overall social budget allows that share to rise without assuming reductions in the absolute amount spent on other areas.

The value‑for-money assessment therefore focusses on whether changing the design or allocation of social spending improves child outcomes relative to an increase in spending alone, while also considering its implications for longer-term spending pressure. A policy is considered to provide relatively strong value for money when its expansion is associated with additional reductions in child poverty while moderating, or at least not increasing, long-term spending pressure. Conversely, a policy provides relatively weak value for money when it weakens the poverty-reducing effect of additional social spending and is associated with greater spending pressure. Some policies may perform well on one dimension but less well on the other, implying a policy trade‑off between stronger child outcomes and lower long-term expenditure.

Two sets of policy changes are simulated with different magnitudes. The first set of simulations is conducted for all policy characteristics and considers a one‑standard-deviation increase in the policy variable, alongside a half-standard-deviation increase in overall social expenditure or GDP per capita. A one‑standard-deviation change reflects the historical variation in the policy characteristic across OECD countries and over time and therefore provides a common metric for comparing policy characteristics measured in different units. The half-standard-deviation changes in social spending and GDP correspond to approximately 4% and 2% of the sample averages of the logged values of total social expenditure per capita and GDP per capita, respectively, and are broadly equivalent to 11 years of average growth in both variables over the sample period.8

A second set of simulations is performed for policy variables that measure the allocation of social expenditure across policy areas. Here, we consider a one‑p.p. increase in the share of social spending allocated to the policy area, alongside the same increase in overall social expenditure or GDP. This second approach allows to compare the relative effectiveness of allocating an equivalent amount of additional resources across different policy areas.

The second set of simulations is intended to inform decisions about how to allocate additional resources across competing policy instruments and, conversely, which areas of spending are less likely to generate large increases in child poverty if spending needs to be reduced. In this sense, the simulations can also help identify policy configurations that may better protect child outcomes in a context of future spending constraints. Policies with a relatively small outcome change – indicating that changing the policy has little effect on the poverty-reducing effectiveness of overall social spending – may be less critical to preserve in the event of spending cuts, although this should not be interpreted as evidence that these policies are unimportant for child well-being.

The simulations do not, however, estimate the effects of a budget-neutral reallocation in which additional spending in one policy area is necessarily financed by equivalent cuts elsewhere. Rather, they provide evidence on how the composition and design of social spending may affect child outcomes and spending pressures conditional on an overall change in social expenditure or economic resources.

Where the interaction between policy design and the core drivers is statistically significant, the estimated value for money also varies with the growth of social expenditure or economic development, giving rise to potentially different policy trade‑offs across countries. Further details on the simulations and their implementation are provided in Box 3.3 “How is the long-run value for money of social policies estimated?”, with additional information on the data, applied imputations and regression results provided in Annex 3.B, Annex 3.C and Annex 3.D.

3.4. The value for money of social policies in reducing child poverty

Copy link to 3.4. The value for money of social policies in reducing child poverty

The empirical analysis studies the core drivers of child poverty, material deprivation and how social expenditures impact them. The results are detailed in the annex tables in Annex 3.D.9 Starting with the influence of core drivers of child outcomes, the results10 confirm that the prevalence of certain demographic trends is associated with a higher risk of children living in poverty and experiencing severe material deprivation:

An increase in the share of children living in single‑parent households is associated with a higher risk of poverty, including severe income poverty (as measured by the poverty line anchored in 2005). Across specifications, a one p.p. increase in the share of single‑parent families is associated with an increase in the child relative poverty rate by around 0.3 p.p., which represents around 5% of the average within-country variation in our sample. Similarly, it is linked to a 0.9‑p.p. increase in the anchored poverty rate, the equivalent of 10% of the average within-country variation in the anchored poverty rate.
Higher rates of teenage pregnancy are linked to greater child relative poverty. Each additional birth per 1 000 women aged 15‑19 is associated with a 0.1‑p.p. increase in the child relative poverty rate.
In contrast, the number of children per household shows no significant relationship with child poverty rates, suggesting that the decrease in the average family size has had no measurable impact on countries’ child poverty levels.
Depth of poverty matters: a one p.p. increase in the mean poverty gap is associated with a 0.7 p.p. rise in child anchored poverty rate and a 0.5 p.p. increase in severe material deprivation, indicating that children are not spared when countries experience a deepening of poverty.

Turning to social expenditures, per capita social spending shows the expected strong negative association with child poverty and material deprivation. A half-standard-deviation increase in the log of social spending per capita – representing the average social expenditure growth over one decade approximately – is associated with an additional reduction of nearly one and a half p.p. in the child relative poverty rate. This magnitude is equivalent to lifting one eighth of income‑poor children out of poverty, on average across the OECD, as measured by 2023 child relative poverty rates. It is also associated with a decline of more than three and half p.p. in the child anchored poverty rate (nearly half the size of today’s average child anchored poverty rates), and over four p.p. in the share of children experiencing severe material deprivation (nearly 70% of children experiencing severe material deprivation).

When analysing the drivers of social spending, the results confirm that increases in GDP per capita are consistently associated with higher levels of per capita social spending.11 A half-standard-deviation increase in the log of GDP per capita – also representing the OECD average of growth over roughly one decade in the past – is linked to an increase of around 9% in social spending per capita. In contrast, within-country variation in the dependency ratio – the population under age 20 and over age 65 as a share of the population aged 20 to 64 – and in life expectancy are not significantly linked to per capita social spending levels.

Figure 3.9 below visualises the value for money simulations for the full set of policy features based on a hypothetical country with average levels of GDP and social expenditure (derived from the regression coefficients in Annex 3.D). The simulated policy increase represents an expansion of the policy by one standard deviation alongside an increase in social spending or GDP by half a standard deviation (around 11 years of social spending and GDP growth, based on historical trends). This allows for a stylised comparison of policy changes across different units.

The quadrants on the bottom left, highlighted in green, display policies with a high value for money. These policies are associated with a higher reduction of the child poverty rate than in the counterfactual scenario in which social expenditure increases without a change in policy design, while also being associated with lower long-term spending pressure – that is, a weaker relationship between GDP per capita growth and per capita social expenditure growth. In contrast, the top right quadrants, coloured red, indicate a low value for money: policies in this area are associated with a smaller decrease or an increase in child poverty or material deprivation than in the counterfactual, whilst at the same time, being associated with greater long-term spending pressure.

The remaining two quadrants comprise social policies that come with trade‑offs between the two dimensions of value for money. Policies in the bottom right quadrant are associated with greater reductions in child poverty or material deprivation but at the cost of higher long-term spending pressure, while policies in the top left quadrant are associated with lower spending pressure but weaker improvements in or worsening child outcomes.

The red horizontal line indicates how much higher child poverty rates would be in the absence of any increase in per capita social expenditure and no change in policy design. The difference between the horizontal axis (0‑line) and the red line represents the estimated poverty-reducing effect of a half-standard-deviation increase in social expenditure, with policy design unchanged and based on the cross-country average policy mix. Policies between the red line and the 0‑line are estimated to attenuate the social spending effectiveness at reducing child poverty compared to the scenario of a spending increase without any policy change, and policies below the 0‑line are estimated to increase the spending effectiveness following an expansion of the policy.

Figure 3.9. The value for money of social policies to reduce child poverty/severe material deprivation, for average levels of per capita social expenditure and GDP per capita
Copy link to Figure 3.9. The value for money of social policies to reduce child poverty/severe material deprivation, for average levels of per capita social expenditure and GDP per capita

A general finding for all measures of child poverty is that across all policy simulations, an increase in social expenditure is always associated to a poverty reduction (as shown by the fact that the resulting change in poverty rates always fall below the red lines). However, the effectiveness of social spending varies substantially depending on policy design, with some policies reducing the effectiveness of social spending in lowering child poverty (those above the horizontal 0‑line) and others enhancing its impact (those below the horizontal 0‑line). The following paragraphs provide further detail on how effectiveness varies across policy settings, focussing on the main categories of policy levers illustrated in the figure.

ECEC enrolment and Active labour market programmes: high value for money

Two dimensions of social spending policies that operate through the labour market show the strongest value for money among the policy areas examined: the enrolment of young children in early childhood and care services and expenditures on active labour market policies. They consistently appear in, or near, the green quadrant at the bottom left of Figure 3.9, indicating that they increase the effectiveness of social expenditure at reducing child poverty while moderating long-term pressures on social spending. A key mechanism is likely higher parental employment, which reduces reliance on income support and lowers expenditure on means-tested benefits. This underscores their role as key levers for promoting parental employment and preventing child poverty, as documented in previous studies (OECD, 2018[33]; Thévenon et al., 2018[8]). More specifically:

Coverage (and take‑up) of Early Childhood Education and Care (ECEC) Services matters more than overall spending levels alone. Expanding ECEC coverage is linked to additional reductions in child income poverty, as well as in severe material deprivation in low-spending contexts. Raising participation in ECEC at ages 0‑2 by one standard deviation (+18 p.p.) is associated with declines in child anchored poverty rates by 1.8 p.p. – thus enhancing social spending effectiveness by 50% – while reducing spending pressures at higher levels of GDP per capita (Figure 3.9, Panel B). Similarly, a one‑standard-deviation increase in ECEC enrolment rates among 3‑5 year‑olds (+18 p.p.) is estimated to double social spending effectiveness at mitigating child relative poverty (an additional ‑1.5 p.p.) and to lower rates of severe material deprivation in low-spending contexts. In high spending contexts on the other hand, further increases in enrolment in pre‑primary education and care appear to have a more limited effect on reducing severe child material deprivation. One likely explanation is that enrolment rates are already very high and services reach the vast majority of children, leaving relatively little scope for additional gains from further expansion. Moreover, the children who remain outside the system are often among the most disadvantaged. For these groups, participation in pre‑primary education alone is unlikely to be sufficient to address the multiple and overlapping forms of disadvantage captured by measures of severe material deprivation. Reducing deprivation among these children is likely to require a broader package of interventions spanning income support, housing, health, and access to other essential services. As a result, in countries where pre‑primary coverage is already near universal, additional investments in other policy areas may yield larger reductions in severe material deprivation than further increases in pre‑primary participation.12
By contrast, the size of ECEC spending in the social budget shows no detectable effect on child poverty. This finding likely reflects that increased spending does not automatically translate into higher coverage, as resources may also be directed towards improving service affordability and quality, or towards extending the number of hours offered to children already enrolled. The findings suggest that ECEC contributes to reducing income poverty primarily when increased investment results in higher enrolment, enabling parents to raise their labour market participation or work intensity, and in turn their earnings. This does not diminish the importance of investing in service quality, which remains essential for children’s well-being and development (see Chapter 4). However, when the objective is to reduce child poverty, it is critical that ECEC investments translate into broader and more equitable access for families.

Expanding access for disadvantaged children therefore represents an important lever for reducing child income poverty. For example, across European countries in 2024, around 49% of children under age 3 from the highest-income third were enrolled in formal childcare, compared with 32% among those from the lowest-income third (see Chapter 2).

ALMPs show poverty-reducing effects across all measures of poverty. A one‑standard-deviation increase in the share of social spending devoted to ALMPs (around two p.p. of total social spending) is associated with reductions of about 0.8 p.p. in child relative poverty, 1.5 p.p. in anchored poverty – which captures deeper forms of income poverty – and 0.9 p.p. in severe material deprivation, while moderating spending pressures by around 2.25% in the value‑for-money simulations (see Figure 3.9). Such a greater policy orientation towards ALMP thus enhances the overall effectiveness of social spending by over 50% for child relative poverty, by around 40% for child anchored poverty, and by 20% for rates of severe material deprivation among children. Although the available data does not allow to single out the particular aspects of ALMP that support parents, the finding suggests that globally, ALMPs are effective in reaching and supporting parents with children on low incomes.13
A greater policy orientation towards ALMPs appears particularly effective at reducing child anchored poverty and severe material deprivation in countries where total social expenditure is low. The effectiveness of allocating higher shares of spending to ALMPs at improving children’s living standards decreases at higher social spending levels (see Annex 1.A). Similarly, it is worth noting that ALMPs’ capacity to ease fiscal pressures depends on the level of economic development: while ALMPs tend to reduce long-run spending in higher-income countries, they may initially add to budgetary pressures in lower-income contexts (see Annex Figure 3.E.1 for a value for money calculation for a poorer hypothetical country).
Overall, the findings suggest that countries allocating a larger share of social expenditure to ALMPs have been better able to reduce child poverty and severe material deprivation. However, the aggregate nature of the data does not allow conclusions about which programme types are most effective. Existing evidence suggests that training and skills-development programmes may generate particularly strong long-term returns by improving earnings potential and labour market matching (see Section 3.5). At the same time, the persistence of in-work poverty across many OECD countries highlights that employment support alone is unlikely to lift all poor families out of poverty. To maximise their impact on children’s living standards, ALMPs therefore need to be complemented by policies that ensure employment translates into adequate living standards for families with children.
Making work pay can further strengthen impacts on living standards. Reducing effective tax rates on labour income for low-income families have the potential to lower the risk of deep income poverty (as captured by “anchored” poverty rates) and to translate into lower exposure to material deprivation but this depends strongly on their design. For example, in the hypothetical case of single parents considered in the analysis, reductions in effective tax rates for single parents increasing their working hours from part-time to full-time employment are associated with lower rates of severe material deprivation. By contrast, lowering the effective taxation of earnings when moving from unemployment to full-time employment at two‑thirds of the average wage shows no significant association with reductions in severe child material deprivation. This suggests that, for single‑parent families, reducing effective tax rates may have a stronger impact on material deprivation when it enables parents who are constrained to part-time work to increase their working hours.
Findings for couple families point in the opposite direction. In the hypothetical case analysed, decreases in effective tax rates when the second parent moves from unemployment into full-time employment are associated with lower rates of severe material deprivation, whereas reductions in effective tax rates linked to transitions from part-time to full-time employment do not show a similar association. This suggests that lowering the effective taxation of earnings for second earners entering employment may play a key role in reducing severe material deprivation in couple families.
Taken together, these findings indicate that the potential of tax-benefit design to contribute to reductions in child poverty depends on several factors: the definition of poverty considered (income poverty or material deprivation),14 the income groups targeted, and the specific labour market transitions involved. Policies that reduce effective tax rates therefore need to take into account the different constraints faced by single‑parent and couple families in moving into employment or increasing their working hours.

Health spending improves children’s living standards and eases fiscal pressure, while other in-kind benefits involve trade‑offs

Broader public services available to the population at large – such as healthcare and social housing – can reduce financial strain on families and indirectly support parental employment and contribute to a greater efficacy of social spending in lowering child income poverty and material deprivation rates. The results of the empirical analysis indicate a high potential of these types of in-kind spending to achieve longer-term improvements in children’s living standards in low-income families, suggesting that such services can reduce barriers to employment for low-income parents in addition to their possible benefits in the primary outcomes that these policies target. However, the findings also point to possible caveats for each measure, drawing attention to the need for effective policy design and implementation if they are to contribute to child poverty reduction and fiscal moderation. More specifically:

Health spending, a major component of in-kind social expenditure, appears particularly effective at reducing deeper forms of child poverty and supporting low-income families. A one‑standard-deviation increase in the share of social spending allocated to health represents a sizeable increase in spending (around 9 p.p. of social expenditure) due to the large shares of social expenditure allocated to healthcare and important sample variations. Consequently, it is associated with a substantial additional reductions in anchored child poverty (‑4.1 p.p., doubling social spending effectiveness compared to the simulation in which the social spending mix remains unchanged) and in child severe material deprivation (‑3.1 p.p., raising social spending effectiveness by 76%), as well as with lower overall spending pressures (‑11%). The relationship with relative child income poverty is more nuanced, however. Higher health spending is associated with stronger reductions in relative child poverty in countries with lower overall levels of social spending, suggesting that health services may reach low-income families particularly effectively in such contexts where overall support remains more limited. By contrast, this association weakens in higher-spending countries, which may reflect shifts in the composition of health spending away from measures that are more strongly targeted at low-income households.
Although the precise mechanisms cannot be directly identified, higher health spending likely improves access to services, especially for low-income families who might otherwise forgo care due to high out-of-pocket costs. Better access to timely and preventive healthcare can help avoid the development of more serious conditions and support parents’ ability to remain in work and earn income. The observed moderation in spending pressures may also reflect employment-enhancing effects, which reduce reliance on cash transfers, including incapacity-related benefits.
In-kind spending on housing, incapacity-related and other social policy areas (e.g. social assistance benefits) contributes to reducing child poverty but is accompanied by higher public expenditure. A one‑standard-deviation increase in the share of social spending devoted to these measures (around 3.5 p.p. of per-capita social spending) is associated with an increase in social spending efficacy to mitigate child relative poverty by 20% (an additional ‑0.3 p.p. compared to the scenario of increasing social expenditure without changing the policy orientation), alongside larger declines in anchored child poverty (‑1.6 p.p., representing an increase in spending efficacy by 40%) and child severe material deprivation (‑2.1 p.p.). At the same time, these increases are linked to higher overall social spending pressures (+5.8%), indicating that such benefits tend to complement rather than substitute for other forms of support. As a result, this type of in-kind provision can be a useful additional tool for assisting children in the poorest households, but careful targeting is essential to limit the additional spending pressures these measures may generate.

Income redistribution effectively reduces less severe forms of child poverty, but entails substantial spending pressure

Income redistribution policies can substantially lower child income poverty, particularly at moderate levels of deprivation, but they are generally less effective in addressing the deepest forms of poverty and tend to place upward pressure on public spending. Specifically:

Two complementary channels drive poverty reduction. Vertical redistribution directs resources toward low-income households regardless of household type, which indirectly benefits children because families with children are overrepresented among the poor. Horizontal redistribution provides additional, child-specific support to households with children – through universal or means-tested benefits or other targeted transfers – helping to offset the costs of raising children and, for low-income families, often lifting disposable income above the poverty threshold. In practice, effective systems combine both approaches by targeting poor households while scaling support for families with children.
Direct income transfers are main drivers of poverty reduction but design matters.
- Tax-break for families. Reductions in personal income taxes granted to families with children show mixed associations with child poverty reduction, depending in part on the overall level of social spending and the objectives these tax measures are designed to serve. A higher share of spending on family-related tax breaks (as a percentage of GDP) appears to strengthen the poverty-reducing impact of social spending in lower-spending countries where social supports are scarce, but is less effective than the average spending mix in higher-spending contexts. Moreover, regardless of the overall level of social expenditure, such tax breaks show no significant association with reductions in child material deprivation, suggesting that they are broadly as effective as other forms of spending. This pattern is consistent with the design of most child-related tax reliefs, which tend to benefit a broad range of families as spending increases – including higher-income households – rather than being tightly targeted at low-income families or at reducing poverty.
- The volume and progressivity of cash benefits distributed to all households reduce child income poverty (including deeper “anchored” poverty). For a given tax structure, a one‑standard-deviation increase in average benefit size (around 6 p.p. of disposable income) is associated with a 1.2 p.p. reduction in child relative poverty, increasing the poverty-reducing effectiveness of social spending by around 80% relative to the baseline policy configuration. In addition, each one‑standard-deviation increase in the share of benefits directed to the bottom 30% of the income distribution – a broader group than those categorised as income poor – is associated with a further 1.1 p.p. reduction in child relative poverty. Similarly, a one‑standard-deviation increase in both, average benefit size and the share that is distributed to the bottom 30% is associated with a 1‑p.p. decline in child anchored poverty rates overall (see specification 8 in Annex 3.D).
- Targeting support towards low-income families helps reduce income poverty, but is less suited to addressing material deprivation. Concentrating income transfers on the poorest households is particularly effective in reducing anchored child poverty: a one‑standard-deviation increase in the share of cash benefits going to the bottom 30% of households is associated with a 3‑p.p. reduction (increasing social spending efficacy by 80%). Conversely, increasing the tax burden on low-income households is linked to higher anchored poverty rates, underscoring the risks posed by regressive financing structures.
  However, stronger targeting towards low-income families – as measured by the share of cash benefits going to the bottom 30% of households – reduces the effectiveness of social expenditure in addressing severe child material deprivation in high-spending contexts. Similarly, Figure 3.9, Panel C indicates that the RED index, which captures the extent of redistribution towards poorer households, lies above to the zero line. This suggests that greater redistribution towards low-income families can come at the cost of effectively addressing severe material deprivation among children.
  As discussed in Section 2.1, many materially deprived children live in households with incomes above standard poverty thresholds and may therefore receive little or no support when benefits are more narrowly concentrated on the lowest-income groups. This points to an important policy trade‑off: while stricter targeting can be effective in reducing income poverty and containing spending pressures, it may be less well suited to addressing material deprivation more broadly.
Horizontal redistribution towards families with children further reduces poverty, but effects vary by family size and spending levels. Greater income support for families with children, relative to otherwise similar childless households, is associated with reductions in child relative poverty of around 1.4 to 1.8 p.p. – equivalent to increasing the poverty-reducing effect of the simulated increase in social spending by around 95‑120%. The association is particularly pronounced in countries with higher levels of social expenditure where the size of transfers is greater. No significant association is found with child deeper (anchored) poverty, possibly because benefit levels are too low or because benefits’ take‑up among the poorest households remains limited (see further discussion in the conclusion).
Reductions in severe material deprivation depend on the overall level of spending: in lower-spending systems, transfers are often too small to reach the most disadvantaged children. As social spending increases, the transfers reaching severely deprived children are likely larger, and are therefore associated with lower rates of severe material deprivation among smaller families. However, for families with three children, the results suggest that the level and design of income transfers remain insufficient to meaningfully reduce severe material deprivation among children, even at relatively high levels of social expenditure. This pattern is consistent with the view that, at lower levels of social expenditure, income support primarily aimed at narrowing living-standard gaps between families with and without children is often too limited in size or insufficiently accessible to reach children experiencing severe economic disadvantage.
As welfare states mature and social spending levels increase, family cash benefits effectively mitigate relative poverty but they become less decisive for material deprivation. Increasing the role of family cash transfers in the policy mix enhances the social spending effectiveness at reducing child relative poverty. A one‑standard-deviation increase in the share of spending allocated to families in cash further reduces child relative poverty by 1.8 p.p. (more than doubling (+120%) the effect of increasing social spending alone), with stronger results at higher levels of social spending where transfers are larger.

At the same time, the effectiveness of a strong policy orientation towards family cash benefits in reducing deprivation diminishes at higher levels of social spending, where complementary policies – such as in-kind services, healthcare, and employment supports – play a larger role. Consistent with this, evidence from the present analysis suggests that in-kind transfers and health-related spending are often more directly linked to reductions in material deprivation than cash family benefits.

Child-related support for older children may have greater scope reduce poverty. Analysing family cash benefits by children’s age suggests that the mixed results of family cash spending are driven by older children: directing a larger proportion of social spending toward children in middle childhood and adolescence is linked to additional reductions in relative child poverty rates. At the same time, greater spending during these periods of childhood is associated with higher rates of severe material deprivation, suggesting that benefits at these ages may be targeted more towards lower-income families or be insufficient in size to cover the material needs of older children. By contrast, increased spending on the early years appears as effective as social expenditure on average across all three poverty outcomes, pointing towards a broadly efficient allocation and distribution of family cash benefits for young children in OECD countries.
However, income transfers tend to raise social spending pressure. While cash transfers are effective in reducing poverty directly and timely, they tend to add more pressure on social spending than other social policies. Income transfers alleviate hardship but appear to generate weaker feedback effects than other forms of social expenditure, which would durably reduce future spending needs. In particular, they do not consistently translate into higher employment or a reduced demand for other public services, meaning that increases in transfers typically add to overall expenditure pressures. These results do not imply that the public returns to spending on family benefits are negative. Instead, they suggest that social income transfers are related to higher social expenditure levels than other child-poverty reducing policies,

At the same time, stronger targeting of benefits towards low-income households can help moderate spending pressures as GDP grows. A one‑standard-deviation increase in the share of social benefits received by the bottom 30% of households is associated with a 2.6% reduction in spending pressure in the value‑for-money simulations (Figure 3.9). However, as highlighted above, the trade‑off is that stricter targeting may reduce the effectiveness of transfers in addressing severe child material deprivation.

Comparing the value for money of equivalent investments (in monetary terms) across social policy domains

As countries seek to improve child outcomes while maintaining fiscal sustainability, an important question is where additional social spending should be directed. Figure 3.10 addresses this issue by comparing the value for money of allocating the same amount of additional spending to different social policy domains. Unlike the previous value for money simulations, which examined increases of one standard deviation in each policy area (representing different absolute amounts of spending), the simulations here compare identical increases in spending across policy domains, thereby providing a more direct assessment of their relative effectiveness in reducing poverty and containing long-term spending pressures.

Specifically, Figure 3.10 shows the estimated additional effects of raising the share of total social spending allocated to a given policy area by one p.p. of social expenditure for a given social spending increase by half a standard deviation (representing around 11 years of expenditure growth on average, based on historical trends). In addition, the figure indicates for each policy domain, to what extent changing the composition of social spending along with a GDP per capita increase by half a standard deviation is associated with lower social spending levels than the rise in GDP alone.

The poverty-reducing impact of reallocating social spending varies substantially across policy areas and poverty measures. The estimates reported in Figure 3.10 confirm that the effectiveness of social spending depends not only on the amount invested, but also on how resources are allocated across policy domains.

Expanding investment in ECEC to increase participation15 emerges as the most effective policy option for reducing child relative income poverty. Allocating an additional one p.p. of total social expenditure to ECEC increases the poverty-reducing effectiveness of social spending by an additional 0.74 p.p. (around 50% higher than the increase in social spending alone). This makes ECEC the strongest performer among all policy areas considered, outperforming both family cash benefits and ALMPs. For instance, allocating an additional p.p. of social expenditure to cash benefits for families as social spending increases is linked to an extra 0.41 p.p. reduction in child relative poverty compared to a proportional increase in social spending across policy areas. The finding is consistent with the important role that affordable childcare plays in supporting parental employment and increasing household incomes.

However, the effectiveness of ECEC investment differs markedly when considering deeper income poverty and severe material deprivation. Its impact on anchored child poverty depends strongly on the level of social expenditures and the maturity of ECEC systems. In countries with relatively low levels of social expenditure, increasing the share of spending devoted to ECEC is associated with a reduction in the effectiveness of social spending in reducing anchored poverty. This suggests that reallocating resources towards childcare may come at the expense of policies that provide more immediate support to families experiencing deep poverty. By contrast, in countries with more developed welfare states and extensive ECEC systems, where enrolment rates and maternal employment are higher (see Annex 1.A), additional investment in ECEC substantially increases the effectiveness of social spending in reducing anchored poverty.

A similar pattern emerges for severe child material deprivation, although in the opposite direction. In lower-spending welfare states, increasing the share of expenditure devoted to ECEC is associated with improvements in the effectiveness of social spending in reducing material deprivation, reflecting the important contribution that access to childcare can make. However, these gains diminish sharply as welfare systems mature. In countries with average or high levels of social expenditure, further expansion of ECEC becomes markedly less effective than other policy options for reducing severe material deprivation. In these contexts, the challenges faced by materially deprived families appear to relate increasingly to housing conditions, health needs and other forms of material hardship that childcare services alone cannot address.

By contrast, health services, other in-kind services and ALMPs emerge as the policy areas most consistently associated with reductions in anchored poverty and severe material deprivation. For anchored poverty, allocating a larger share of social expenditure to ALMPs generates the largest increase in spending effectiveness, followed closely by health spending and other in-kind services such as housing, incapacity-related benefits and social services. Allocating an additional one p.p. of the social spending mix to ALMPs when spending expands by half a standard deviation enhances the effectiveness of the social spending increase by an additional ‑0.6 p.p. (17% higher than the effect of increasing social spending proportionally across policy domains). In comparison, an additional p.p. of social expenditure directed to in-kind spending on incapacity-related, housing and other social is linked to a further reduction in child anchored poverty by 0.5 p.p. (corresponding to a 12%-increase in overall spending effectiveness).

For severe material deprivation, other in-kind services produce the strongest gains (an additional ‑0.6 p.p., enhancing spending effectiveness by 14%), while shifts towards ALMPs and health spending also improve spending effectiveness (by an additional ‑0.4 and ‑0.3 p.p., respectively). These findings suggest that policies addressing employment barriers, health needs and families’ material living conditions are particularly important for tackling deeper and more persistent forms of disadvantage. In contrast, a greater use of family cash benefits in the policy mix as spending increases is linked to reductions in social spending effectiveness (+0.4 p.p., lowering spending effectiveness by 9%).

Moreover, similar patterns emerge for long-term spending sustainability. Greater investment in expanding ECEC coverage is associated with lower long-term spending pressures in more mature welfare states (1.9% lower social spending – a reduction in additional social spending by one fifth compared to the projected expenditure growth associated with an increase in GDP and no change in the allocation of social spending), where higher parental employment and stronger labour market attachment increasingly offset the costs of service provision. In less developed economies, however, expanding ECEC may initially increase social expenditure further as systems are developed and other policy supports remain scarcer (see Annex Figure 3.E.1). Health spending and active labour market programmes (ALMPs) also help contain long-term spending growth (by an additional 1.2% and 1%, respectively), whereas allocating a larger share of social expenditure to family cash benefits and some other in-kind services is associated with higher future spending (around 1.4% and 1.7%, respectively).

Overall, the comparison across poverty measures highlights an important distinction. Investing a fixed amount of additional spending in ECEC to increase participation rates delivers the largest returns over time for reducing child relative income poverty as well as for child anchored poverty when coverage and the overall level of social expenditure are high. These gains in spending effectiveness are achieved through ECEC’s effects on parental employment and earnings. However, when the objective is to reduce severe material deprivation, health, housing, other in-kind services and ALMPs offer better value for money. As welfare systems develop, the marginal contribution of additional ECEC expansion declines for combating severe material deprivation, while the relative importance of policies that directly address families’ needs of in-kind services increases.

Figure 3.10. Allocations across policy areas shape spending effectiveness at reducing child poverty and social expenditure levels
Copy link to Figure 3.10. Allocations across policy areas shape spending effectiveness at reducing child poverty and social expenditure levels

Estimated effects of a one‑p.p. increase in the share of social expenditure allocated to different policy domains on child poverty outcomes and total social spending levels

3.5. Beyond how much we spend: getting policy design right

Copy link to 3.5. Beyond how much we spend: getting policy design right

The findings in this chapter demonstrate the value of combining employment-oriented social policies with other types of in-kind support and income transfers to reduce child poverty. These policy instruments are complementary in that they address different needs of families with children – both in preventing poverty and in supporting exits from it. Each area of support has an independent poverty-reducing effect when considered on its own, but the evidence also shows that the impact of social policy measures is often stronger when different types of support are implemented together. Specifically, the findings from this chapter are consistent with a broad body of evidence showing that a balanced policy strategy – one that combines measures to strengthen employment opportunities and market incomes with redistribution through cash transfers – is the most effective approach to reducing child poverty (Bäckman and Ferrarini, 2010[64]; Whiteford and Adema, 2007[65]; OECD, 2018[33]).

A key contribution of this chapter is to identify specific policy design features that shape effectiveness in reducing child poverty and that exert differentiated pressures on social spending. The analysis also highlights that policy effectiveness is not static, but evolves dynamically as welfare states mature and social expenditure increases. This section situates the empirical findings within the broader evidence on the impact of social policies on child poverty. It also highlights key implementation challenges for designing policy packages that can deliver the greatest value for money – that is, policies capable of enhancing poverty reduction outcomes while containing pressures on government social spending, in line with the patterns identified in the preceding analysis.

Different types of social policy address distinct dimensions of families’ needs and operate over different time horizons, making their combination an important determinant of policy effectiveness. Pro‑employment measures aim to secure stable earnings and address long-term risks but often take time to produce effects (Card, Kluve and Weber, 2017[57]). Cash transfers, by contrast, respond to immediate and temporary needs while also supporting households’ capacity to plan and invest over the longer term. In-kind services play a dual role: they improve living conditions and access to essential services, and they can facilitate labour market participation, particularly for disadvantaged groups. Effective policy design therefore requires combining these instruments in a coherent way, rather than prioritising any single approach in isolation as illustrated in the rest of this section that distinguishes between packages of social policies.

Boosting the poverty-reducing impact of pro‑employment policies

The results presented in the previous section show that pro‑employment policies are particularly effective in lifting children out of the most severe forms of poverty and that stronger labour market attachment can help contain future social spending pressures. However, the analysis also shows that the relationship between different pro‑employment policy levers and child poverty reduction varies. Each policy instrument faces indeed specific challenges in maximising its effectiveness as a tool for poverty reduction and there are specific barriers that could be removed:

Enhancing participation in ECEC services for children from low-income families

ECEC provision alone is not sufficient to address the most severe forms of child disadvantage. This also reflects that, in many countries, ECEC services are not primarily accessed by low-income families (see Chapter 2). Expanding access and participation among children from low-income and materially deprived households therefore remains an important policy challenge for strengthening the poverty-reducing potential of ECEC systems. Addressing this challenge requires overcoming a set of interrelated barriers that drive a persistent “Matthew effect”, whereby advantaged families benefit first. On the supply side, limited capacity, long waiting lists, and uneven geographic provision mean that better-connected or higher-income parents who can commute, pay higher fees, or navigate application systems more effectively, are more likely to secure places (OECD, 2025[70]). Affordability also remains a major constraint, as fee structures and poorly targeted subsidies often leave low-income families facing higher effective costs or complex administrative hurdles (OECD, 2020[71]). Demand-side factors also play a role: differences in employment patterns, access to information, language skills, social networks, and the ability to manage irregular hours can significantly reduce uptake among disadvantaged families, beyond differences in parental preferences alone (Wood, Neels and Maes, 2023[72]). Finally, policy design and sequencing matter: expansions that prioritise quality or capacity without simultaneously addressing fees, accessibility, and complementary supports tend to reinforce existing inequalities in access rather than reduce them (Van Lancker, 2023[73]). Child poverty reduction appears to be particularly strong when ALMPs are implemented alongside expanded access to early childhood education and care (ECEC) services. By helping to ease childcare constraints, ECEC enables parents to participate more sustainably in the labour market and to benefit more fully from employment support measures. This complementary effect is consistent with previous findings by Nieuwenhuis (2022[74]) who found that ALMPs are associated with reductions in women’s unemployment rates, while greater availability of ECEC is linked to lower inactivity rates among women.

Designing effective tax rates to make work pay

The findings from this chapter suggest that the responsiveness of tax and transfer systems to labour supply decisions significantly shapes the effectiveness of this support, highlighting the importance of strengthening the financial returns to work not only at the point of labour market entry but also when working parents increase their labour supply. Even where clear reductions in income poverty are not observed, higher take‑home pay can reduce material deprivation, which – as shown earlier in the chapter – affects many families whose incomes remain above the poverty line. Furthermore, these increases in take‑home pay do not appear to add to social spending pressures, which may highlight that they are well targeted towards lower income families and that they are likely to generate a lower overall reliance on social benefits by incentivising greater labour market activity. More broadly, strengthening in-work incomes can help reduce in-work poverty while containing pressure on social spending by complementing earned income. Importantly, the available evidence that income supplements must be strictly conditioned on employment to avoid labour-supply disincentives among low-income families remains limited (Box 3.5).

Box 3.5. Reconciling work incentives and income support for low-income families: is there a trade‑off?

Copy link to Box 3.5. Reconciling work incentives and income support for low-income families: is there a trade‑off?

A frequently cited concern in the design of income support for families with children is that lump-sum cash transfers may reduce parental labour supply through income effects, by lowering incentives to work or shifting time away from paid employment. In response, many countries have chosen to link part of family support to employment, through in-work tax credits or employment-conditional child allowances, with the aim of simultaneously supporting family incomes and strengthening labour-market attachment among low-income parents. In practice, evidence from OECD countries indicates that such work-conditioned instruments contribute to substantial reductions in child poverty – particularly deep child poverty – while also increasing employment among disadvantaged groups (Michelmore, 2025[75]). For example, evaluations of the Earned Income Tax Credit (EITC) and Child Tax Credit (CTC) in the United States show that these work-linked refundable credits lifted millions of children out of poverty in the early 2010s; and research indicates that incorporating the employment-boosting effect of these credits nearly doubles their measured anti-poverty impact for low-income single mothers when employment gains are taken into account (Marr et al., 2015[76]). Beyond short-term income effects, exposure to the EITC during childhood has been shown to generate lasting benefits, including reduced reliance on public assistance in adulthood and a lower probability of being in or near poverty by around 7 p.p. (McInnis et al., 2023[77]).

At the same time, recent quasi‑experimental evidence suggests that removing work requirements from child-related tax benefits has very limited adverse effects on labour supply, calling into question the necessity of strict work tests to sustain employment. A 2024 study using administrative data from California finds that extending a child tax credit to non-working parents had negligible effects on maternal labour-force participation, with estimated changes statistically indistinguishable from zero and confidence intervals ruling out even moderate employment declines (Goldin et al., 2024[78]). Similarly, evaluations of the temporary 2021 expansion of the US Child Tax Credit – which substantially relaxed work conditions while increasing benefit size – find small, inconsistent, and statistically insignificant effects on parental employment, reinforcing the conclusion that unconditional or weakly conditional child benefits do not meaningfully disincentivise labour supply among low-income parents (Ananat et al., 2024[79]). Comparable results are reported for Canada, where the introduction and expansion of the work-unconditional Canada Child Benefit had little to no detectable impact on the labour supply of either single or partnered mothers (Thévenon, Clarke and Simard-Duplain, 2024[80]).

More broadly, while activation requirements can contribute to more efficient public spending and help limit long-term benefit dependency, they also entail risks. If poorly designed or overly stringent, such requirements may push vulnerable individuals into low-wage or poor-quality jobs, constrain long-term earnings prospects, and place disproportionate burdens on parents facing multiple barriers to employment (Mroczka, Pacifico and Chen, 2025[81]).

Taken together, this evidence suggests that countries seeking to strengthen support for the poorest families could consider expanding unconditional child transfers. The absence of work conditions is particularly relevant for families experiencing deep poverty, who are often weakly or insecurely attached to the labour market and have limited access to other income sources (Koebel and Stabile, 2024[82]).

Strengthening the role of active labour market policies for parents

Increases in the share of social spending devoted to ALMPs are associated with lower child income poverty rates, including its more severe forms, as well as exposure to severe material deprivation. Nevertheless, further research is needed to consolidate this finding and to identify which specific components of ALMPs are most effective in improving parents’ labour market outcomes in ways that sustainably reduce children’s exposure to income poverty and material hardship.16 In addition, while ALMPs appear effective in tackling poverty, including its most severe forms, an important challenge for further strengthening their impact lies in addressing the rise in in-work poverty observed in many countries, which limits the ability of employment alone to lift families with children out of poverty.

The poverty-reducing potential of ALMPs ultimately depends indeed on the quality and accessibility of the jobs created in the labour market. Their effectiveness hinges on whether economic growth generates sufficient employment opportunities for socio‑economically disadvantaged parents, with wages high enough to lift families out of poverty – conditions that are not always met. Evidence shows that job creation can be unevenly distributed, often concentrated among adults with higher levels of education and stronger attachment to the labour market. In several countries, employment growth has coincided with an expansion of low-quality, low-paid work (OECD, 2015[83]; Corluy and Vandenbroucke, 2017[84]), see Chapter 2. As a result, rising employment does not necessarily translate into significant movements of children out of poverty if in-work poverty increases, particularly among parents in low-paid occupations. For example, Bárcena-Martín (2018[85]) found that across European countries, a one‑p.p. increase in the working-poor rate raises a child’s odds of being income poor by 5.6%, whereas a one‑p.p. increase in parental employment reduces the odds by only 2.4%. As highlighted earlier in this chapter, in the average OECD country around one in ten families with at least one working adult remains income poor, confirming that parental employment does not automatically guarantee protection from poverty.

These findings underscore the importance of strengthening ALMPs and work-family support policies so that parents have opportunities to upskill, progress in their careers, and increase their earning capacity. Ensuring that employment provides living wages, namely wages that are sufficient for families with children to live above the poverty line remains a central challenge (Balestra, Hirsch and Vaughan-Whitehead, 2023[86]). This is particularly important given the rise in in-work poverty observed in many countries. As highlighted in Chapter 2, while minimum wage policies have contributed to protect households from income poverty – including severe income poverty – income transfers to poor households have declined in several countries. Taken together, these results highlight the need to complement employment policies with well-designed income support measures that supplement earnings. Such measures can help ensure that families with children have sufficient resources to meet their needs while preserving the financial incentives associated with work.

Box 3.6. What does the evidence say about the effectiveness of active labour market policies?

Copy link to Box 3.6. What does the evidence say about the effectiveness of active labour market policies?

The evidence base on the effectiveness of active labour market policies (ALMPs) is strongest for employment outcomes. However, impacts are not immediate: evaluations show that average employment effects are close to zero in the short term but become positive two to three years after programme completion (Card, Kluve and Weber, 2017[57]).

The effectiveness of ALMPs varies by programme type and over time. “Work-first” approaches, which emphasise rapid entry into employment through job-search support or job search monitoring tend to generate stable but relatively low impacts over time. “Human capital” approaches such as training programmes often show limited or even negative short-term effects – reflecting time spent out of work – but yield stronger positive impacts in the medium to longer term, especially when the training is aligned with labour market needs. Private sector employment programmes, including wage subsidies and employment incentives, stand out as consistently providing substantial, positive long-run employment effects. By reducing labour costs or supplementing earnings, they aim to facilitate entry into employment and increase disposable income. Public sector employment programmes, including direct job creation schemes, can provide small gains in short-term employment but they generally show negligible or negative effects on long-term employment outcomes.

Programme effectiveness also differs across population groups. Women and long-term unemployed individuals tend to benefit more from ALMPs, while impacts for youth and older workers are typically less positive. There is also evidence that matching programme types to participant characteristics matters: work-first approaches appear more effective for disadvantaged groups, whereas training programmes yield stronger results for the long-term unemployed (Card, Kluve and Weber, 2017[57]).

ALMPs also interact with other policy areas. For example, both ALMPs and early childhood education and care (ECEC) policies contribute to higher female employment, albeit through different channels – ALMPs primarily reduce unemployment, while ECEC reduces inactivity (Nieuwenhuis, 2022[74]). However, evidence suggests diminishing marginal returns when investment in one policy area is already high, indicating some degree of substitution rather than complementarity.

Country-level evidence confirms these general findings. A recent evaluation of ALMP programmes in Slovenia shows that a wide range of interventions – including classroom training, on-the‑job training, and wage subsidies – significantly improve participants’ employment and earnings outcomes (OECD, 2026[87]). These effects are sustained over time, with positive impacts persisting for at least four years after programme entry. Participants are also less likely to exit the labour force or emigrate, and they experience cumulative net income gains compared with similar non-participants. When broader economic effects are considered, all programme types generate net social gains within a four‑year horizon. Evidence on wider social outcomes is more limited. Some programmes, particularly public works, show small positive effects on health and social outcomes for specific groups, but overall differences between participants and non-participants remain modest.

Despite strong evidence on labour market impacts, there is a notable gap in research on the effects of ALMPs on child outcomes. Existing studies are scarce and inconclusive. For example, one study of an employment programme in Madrid finds no significant reduction in material hardship compared with a life‑skills intervention (Ayala and Rodríguez, 2013[88]). A small number of earlier studies from the United States suggest that employment may reduce hardship among single mothers receiving welfare (Danziger et al., 2000[89]; 2002[90]), but their descriptive nature and historical context limit the strength of these conclusions (Grant and Prickett, 2023[91]). Overall, the current evidence base remains insufficient to draw firm, causal conclusions about the impact of ALMPs on child well-being or material hardship, although the results from the correlational analysis in this chapter point in this direction.

Investing in health and housing help reduce child material deprivation

While the present analysis indicates that health spending is associated with reductions in child poverty, an important challenge for policy is to identify more precisely which components of health expenditure – beyond broad spending categories – generate the largest poverty-reducing effects. To a lesser extent, a greater policy allocation towards in-kind spending that helps households with housing costs, incapacity-related benefits and spending in other social policy areas (e.g. social assistance) is also associated with lower rates of child income poverty and material deprivation, although without the same moderating effect on overall social expenditure dynamics. This pattern is consistent with evidence showing that non-child-specific benefits – such as housing support – can, in some contexts, be particularly effective in alleviating material deprivation among children (Bárcena-Martín et al., 2017[92]). It should be noted, however, that the data used in this analysis do not capture all forms of housing support. Other instruments, notably social housing, may also play an important role in reducing child poverty and material deprivation, particularly among highly vulnerable groups such as families experiencing homelessness or severe housing insecurity.

Enhancing the effectiveness of income transfers in addressing child poverty

Income transfers through social cash benefits and tax reductions are a key policy lever for reducing child income poverty. They provide a safety net for families unable to work, complement earnings to protect against in-work poverty, and cushion temporary income shocks. By providing greater income stability, cash benefits can also help families plan expenditures and invest in children’s health, education and development. Universal child benefits can provide a stable foundation for child income protection, complemented by targeted cash or in-kind support where needed (ILO, UNICEF and Learning for Well-Being Institute, 2024[93]; Waldfogel, 2025[94]; Atkinson, 2011[95]).

Evidence shows that transfers are particularly effective in reducing child poverty among households with low work intensity, multiple children and single‑parent families (Chzhen, 2017[96]; Thévenon et al., 2018[8]) as well as single‑parent families (Van Lancker, Ghysels and Cantillon, 2015[97]; Bäckman and Ferrarini, 2010[64]). However, living standards among low-income families declined in several OECD countries between the mid‑2000s and mid‑2010s, particularly among children in very low-income households (OECD, 2018[33]; Thévenon et al., 2018[8]). Chapter 2 also documented a broader weakening of redistribution towards lower-income families, reflecting changes in benefit design and a shift in social spending away from income transfers.

The regression analysis presented earlier finds that, on average, allocating a larger share of social expenditure to family cash benefits is associated with greater reductions in child relative income poverty. Their effectiveness in reducing anchored poverty, however, depends critically on targeting benefits towards lower-income households. The relationship with severe material deprivation is weaker and varies with overall spending levels: while allocating a larger share of social expenditure to family cash benefits and directing a greater share of benefits to the bottom 30% of the income distribution are both significantly associated with lower rates of severe material deprivation when total social expenditure is low (see Annex 1.A), these policies appear less effective than other measures and are linked to reduced overall spending effectiveness in countries with average or larger social budgets. This does not imply that cash support cannot alleviate material hardship: micro-level evidence consistently shows that increases in household income can reduce material deprivation, although effects vary with programme design (see Box 3.7). Moreover, some forms of family-related support, such as supplements embedded in unemployment or housing benefits, are not classified as family cash benefits and may therefore be understated in aggregate spending measures.

Overall, the findings highlight that the effectiveness of cash transfers depends not only on how much is spent, but also on how benefits are designed and targeted. The remainder of this chapter discusses core design features of cash benefits that enhance their effectiveness in mitigating diverse child poverty risks across diverse family types.

Box 3.7. Effects of income support on material hardship and child outcomes

Copy link to Box 3.7. Effects of income support on material hardship and child outcomes

Recent reviews of the literature examining policy-induced variation in household monetary resources – by Page (2024[16]), Waldfogel (2025[94]), Grant and Prickett (2023[91]) and (European Commission: Directorate-General for Employment, 2026[98]) – provide robust evidence on how different forms of income support affect material hardship and child outcomes. This evidence spans a wide range of instruments, including tax credits, direct cash transfers, unemployment and disability benefits, and near-cash transfers such as food and housing assistance.

Tax credits and direct cash transfers

Tax credits, particularly refundable ones, play a central role in delivering income support to families (Moffitt, 2015[99]). Evidence from Canada shows that increases in child benefits – such as the Canada Child Tax Benefit and the National Child Benefit Supplement – improve children’s mental health and educational outcomes, with stronger effects among disadvantaged families (Milligan and Stabile, 2011[100]).

In the United States, evidence on the Earned Income Tax Credit (EITC) points to generally positive, though sometimes uneven, effects on material hardship (Grant and Prickett, 2023[91]). Using variation in state‑level supplements, Davison et al. (2018[101]) find that individuals in states without an EITC were more than 15% more likely to report hardship around the tax filing period (p < .001), with reductions observed across everyday medical, and financial hardship domains in EITC states. This is consistent with evidence that lump-sum EITC payments – typically received annually – enable households to meet larger or “lumpy” expenditures (such as medical expenses or durable goods), reduce debt, and increase savings (Barrow and McGranahan, 2000[102]; Goodman-Bacon and McGranahan, 2008[103]; Gao, Kaushal and Waldfogel, 2009[104]), thereby alleviating current hardship while also buffering against future shocks. Quasi‑experimental studies further indicate that EITC expansions can reduce specific forms of hardship, particularly medical hardship (Kondratjeva et al., 2021[105]), although effects on housing or food hardship are less consistent, likely reflecting the size of transfers relative to these costs. At the same time, a randomised controlled trial of an expanded EITC for low-income workers without children finds no significant reduction in severe material hardship over three years (Miller et al., 2018[106]), despite improvements in income and employment. This suggests that impacts depend on benefit size, target population, and how hardship is measured. Overall, the literature indicates that the EITC reduces material hardship – especially among families with children – through both direct income effects and increased employment (Hoynes, 2019[107]; Hoynes and Rothstein, 2017[108]), although its effectiveness varies across hardship domains and programme design.

Beyond material hardship, increases in income associated with the EITC are also linked to improvements in children’s cognitive outcomes, health at birth, and longer-term educational and labour market outcomes (Page, 2024[16]). These effects reflect both higher household resources and changes in parental labour supply, particularly among single mothers. Evidence from welfare‑to-work experiments further suggests that increases in income – rather than employment alone – are key to improving child outcomes (Duncan, Morris and Rodrigues, 2011[109]). Similarly, evidence on the Child Tax Credit shows consistent reductions in material hardship, with particularly strong effects among families in poverty (Pilkauskas et al., 2022[110]; Collyer Sophie, Wimer and Harris, 2022[111]). For example, Collyer et al. (2022[111]) find that monthly CTC payments reduced the likelihood of experiencing hardship by around 9.5%, with larger effects – up to 18% – among families living in poverty.

Across the EU-27, EU tax-benefit microsimulation model (EUROMOD) simulations show that in-cash child benefits substantially reduce child poverty and its severity, lowering the child at-risk-of-poverty rate (AROP) by an estimated 10.6 percentage points on average and the child poverty gap by 5.7 percentage points (European Commission: Directorate-General for Employment, 2026[98]). Their impact on material deprivation is smaller, reducing the share of children deprived of at least one item by around 3 percentage points on average.

Unemployment and disability benefits

Unemployment and disability benefits also play an important role in cushioning income shocks and addressing material hardship. Evidence from the expansion of unemployment insurance (UI) during the COVID‑19 pandemic in the United States shows that increased benefit size helped stabilise material hardship among recipients, while hardship increased among non-recipients (Grant and Prickett, 2023[91]). Effects were particularly visible in reduced food insecurity and unmet medical needs. By contrast, evidence on disability benefits indicates that recipients often experience higher levels of hardship than non-recipients, largely reflecting underlying health conditions and greater needs rather than programme ineffectiveness (Schmidt and Danziger, 2012[112]). This highlights the importance of accounting for selection into programmes when interpreting estimated impacts.

Near-cash transfers (in-kind benefits)

Near-cash transfers, such as food and housing assistance, also contribute to reducing hardship and improving child outcomes, although typically to a lesser extent than direct cash transfers. Evidence on the Supplemental Nutrition Assistance Program (SNAP) shows that it reduces both food insecurity and non-food hardship, including housing and utility hardship (Shaefer and Gutierrez, 2013[113]; Han, 2016[114]). Studies exploiting variation in the historical rollout of the programme find that early-life exposure improves birth outcomes, long-term health, and economic self-sufficiency, with particularly strong effects when exposure occurs during early childhood (Hoynes, Schanzenbach and Almond, 2016[115]). Housing assistance can similarly reduce financial pressure by lowering rent burdens, although evidence on its effects on children’s outcomes is more mixed. For example, a large housing voucher lottery in Chicago finds limited impacts on children’s long-term outcomes, possibly reflecting the specific characteristics of the target population and programme design (Jacob, Kapustin and Ludwig, 2015[116]).

Other cash transfer programmes

Evidence from other cash transfer programmes, including those with weaker links to employment, confirms the importance of income effects. Historical evidence from the Mothers’ Pension programme in the United States shows long-term gains in health, education, and earnings among children who received transfers (Aizer et al., 2016[117]). Evidence from negative income tax experiments conducted in the 1970s point to improvements in child health and educational outcomes, although effects on long-term outcomes appear limited when income gains are not sustained over time (Price and Song, 2018[118]). More recent evidence on Supplemental Security Income (SSI) suggests that while transfers are substantial, impacts on early child development are less clear, reflecting both small sample sizes and the complex needs of recipient families (Guldi et al., 2022[119]).

Improving the adequacy of benefits for different family types, and sustaining it over time

Programme effectiveness is contingent on the alignment of support with families’ needs and benefit adequacy. This implies ensuring that benefit levels reflect household circumstances, including family structure and size, as children in single‑parent and large families face disproportionately high risks of poverty and deprivation (see Section 3.1). It also implies maintaining support across key stages of childhood and into the transition to adulthood, when poverty risks often remain elevated, despite wide cross-country differences in the duration and design of child benefits (Box 3.8). In addition, effective alignment must take into account increasingly complex family arrangements, such as reconstituted families, where parents’ financial responsibilities may extend beyond the household (Rapp and Thévenon, 2025[29]; Miho and Thévenon, 2020[120]). Families with disabilities also require additional and tailored support to compensate for higher costs and greater poverty risks related to the disability. Given the multifaceted drivers of material hardship – including low wages, insufficient social protection, and broader structural factors – combinations of programmes can be particularly effective. Several studies find that participation in multiple forms of support is associated with larger reductions in hardship (Grant and Prickett, 2023[91]; Guio et al., 2022[121]).

This chapter’s regression analysis points to potential efficiency gains in the design of family cash benefits for older children specifically. The results suggest that allocating a greater share of child-related support to families with preschool children does not alter the overall social spending effectiveness for any of the three poverty measures. The allocation and distribution of family cash benefits during the early ages therefore appears broadly adequate in OECD countries. Instead, directing greater spending shares to family benefits during middle childhood and adolescence is associated with further reductions in children’s exposure to relative income poverty on the one hand but with a reduced overall spending effectiveness with respect to severe material deprivation on the other hand. This could reflect that child benefits at older ages tend to be more targeted towards income‑poor families thus excluding households experiencing severe material deprivation above the income poverty threshold. It might also point to greater needs for support and thus a lower relative effectiveness of cash benefits to address material deprivation among older children and adolescents, whose deprivation levels are often comparable to, or higher than, those of younger children (see Figure 3.8).

Box 3.8. Main characteristics of family benefits

Copy link to Box 3.8. Main characteristics of family benefits

Families’ needs also evolve over the child life course. As a result, many countries provide additional or differentiated support depending on children’s ages – for example, by targeting some benefits to families with pre‑school children or offering supplementary support when children reach school age. Such age‑sensitive designs aim to reflect variations in costs, caregiving demands, and labour-market constraints faced by parents at different stages. There are also important cross-country differences in the definition of resources used for means-testing. While most systems rely primarily on earned and transfer income, some countries additionally apply asset tests, reducing or withdrawing benefits once financial assets exceed specified thresholds (OECD, 2024[122]; OECD, n.d.[123]).

Eligibility criteria for child benefits and other income supports play a critical role in determining their effectiveness in reducing child poverty and material deprivation. Key design features include the age range over which benefits are paid and the conditions attached to receipt, such as means tests and/or work requirements.

In most OECD countries, at least one child or family benefit is payable up to age 17. However, in several countries continued eligibility beyond compulsory schooling is conditional on children being enrolled in education and classified as dependent on parental income (OECD, 2024[122]; OECD, n.d.[123]). In some cases, benefits are payable above age 18 only if young people reside in the parental household and remain financially dependent. While a few countries provide higher benefit amounts as children grow older, this age‑profiling is the exception rather than the rule.

This design contrasts with observed poverty risks and costs over the child life course. In many European countries, child poverty rates among adolescents aged 11 and over are at least as high as, and often higher than, those for younger children (see Chapter 2). Moreover, detailed evidence on the cost of raising children shows that adolescents and dependent young adults aged 14‑24 generate expenses that are at least as high as those of children under age 5 (Rapp and Thévenon, 2025[29]). In several countries – including Germany, Ireland, Spain and the Netherlands – costs increase with age in couple households, while in France this pattern is observed among single‑parent families.

These findings indicate that financial pressures on families do not diminish as children grow older. Aligning benefit design more closely with observed costs and poverty risks therefore implies ensuring adequate and continuous income support for families with adolescent children, rather than concentrating support primarily in the early years.

Source: OECD (n.d.[124]), “How do countries calculate tax liabilities and social benefit entitlements?”, https://www.oecd.org/en/topics/sub-issues/income-support-redistribution-and-work-incentives/how-do-countries-calculate-tax-liabilities-and-social-benefit-entitlements.html and OECD (n.d.[125]), OECD Family Database, https://www.oecd.org/en/data/datasets/oecd-family-database.html.

Moreover, maintaining benefit adequacy over time is essential. Beyond establishing appropriate payment levels at the outset, financial supports must be periodically adjusted to account for increases in the cost of living. In many systems, child benefits are not automatically indexed to inflation, leading to a gradual erosion of their real value and weakening their capacity to protect families facing deep poverty (Box 3.9).

Box 3.9. Indexation of child and family benefits in OECD countries: practices and implications

Copy link to Box 3.9. Indexation of child and family benefits in OECD countries: practices and implications

Across OECD countries, there is no single standard approach to the indexation of social benefits, including child and family benefits. Instead, countries rely on a mix of automatic inflation indexation, discretionary uprating, and ad hoc adjustments, leading to substantial cross-country variation in how well benefit levels are protected against rising prices over time (see Annex 3.F for a comparison of selected countries).

In a first group of countries, child and family benefits are automatically indexed to inflation or a cost-of-living index, helping to preserve their real value. This is the case, for example, in Belgium, where child benefits are indexed to a price index (the “health index”) through automatic mechanisms – e.g. automatic increases tied to a price index when a pivot point is reached, and in Canada, where the Canada Child Benefit is updated annually based on the Consumer Price Index. Luxembourg also operates an indexation mechanism linked to its general indexation system. Automatic indexation reduces the risk of real-term erosion and ensures predictable protection of purchasing power, but it implies higher and more responsive fiscal costs during inflationary periods (OECD, n.d.[124]; 2023[126]).

In many other OECD countries, however, indexation is partial or discretionary rather than automatic. France, for instance, typically revalues family benefits annually based on inflation forecasts through the adjustment of the Base mensuelle de calcul des allocations familiales, but these increases are set by decree and may lag behind actual inflation (HCFEA, 2023[127]). Germany, Austria, Italy, Spain and the United Kingdom rely largely on periodic legislative or budgetary decisions to uprate child benefits, with no strict rule tying increases mechanically to inflation. As a result, benefit levels in these countries have often lost real value over time, particularly during periods of high inflation, unless governments introduced exceptional or one‑off increases (OECD, 2023[126]; n.d.[124]).

Overall, OECD evidence shows that while many countries adjust child and family benefits with reference to inflation, automatic and full indexation remains far from universal. Where indexation is discretionary, child benefits are more exposed to political and fiscal constraints, increasing the risk of a gradual erosion in their real value. This has important implications for the effectiveness of family policies in protecting living standards and reducing child poverty, particularly during periods of high inflation.

Indexation to inflation also often implies that benefits tend to grow more slowly than nominal wages, leading to a declining share of GDP over time. In France, for example, average wages per capita have increased by around 0.5 to 1 p.p. faster than consumer prices each year since 2012. As a result, price indexation has a twofold effect: first, eligibility for means-tested benefits gradually erodes as incomes rise faster than thresholds; second, benefit amounts become less dynamic than average nominal wages. While this mechanism dampens expenditure growth, it may also create fiscal space to rebalance priorities or finance policy reforms (IGAS, 2021[128]).

Beyond child benefits, the indexation of personal income tax (PIT) parameters, social security contributions (SSCs), and other social benefits also plays a critical role in determining how redistribution adjusts to changes in the cost of living (OECD, 2023[126]). In several countries, inflation-adjustment rules differ across these three policy levers. As of 2022, fewer than half of OECD countries (17) automatically indexed PIT parameters, while adjustments were discretionary in 20‑one countries. Most countries indexed SSCs, and around half automatically indexed social benefits. Overall, 12 OECD countries applied automatic indexation to PIT, SSCs, and benefits alike, whereas ten relied entirely on discretionary adjustments across all three categories.

Targeting income support: Trade‑offs between spending control and reducing material deprivation

There is considerable diversity in how income support systems for families are designed, with marked differences across countries in the degree of income targeting embedded in these schemes and in the size of support provided to low-income families. In some cases – such as Canada, Ireland and Poland – benefits are strongly concentrated at the bottom of the income distribution, with very high support for low-income families followed by a steep taper as income rises (Koebel and Stabile, 2023[129]; 2024[82]). In others, including Denmark, France and Germany, benefit levels for lower-income families are comparatively lower and decline more gradually with income, resulting in a broader distribution of support across the income distribution.

The findings in this chapter extend earlier evidence by highlighting the trade‑offs associated with targeting income support towards low-income households. On the one hand, stronger targeting of social benefits can help contain spending pressures linked to economic development and the growing needs of ageing populations. This likely reflects the fact that higher income transfers do not typically generate feedback effects that permanently reduce future social spending needs. While transfers can alleviate poverty and material hardship, they may not substantially reduce demand for other forms of public spending, such as healthcare, education, housing support or pensions, nor do they necessarily translate into higher employment rates that would ease pressure on overall social expenditure. At the same time, increasing the share of cash benefits directed to the bottom 30% of the income distribution – regardless of the average size of the total benefits – is associated with greater effectiveness in reducing both relative child income poverty and deeper poverty, as captured by poverty rates anchored in 2005.

However, while tighter targeting of income transfers towards low-income households can be effective in reducing income poverty, it is not associated with reductions in children’s exposure to severe material deprivation in all contexts. When spending levels are high, targeting income benefits appears less effective than other social policies at mitigating child severe material deprivation. This finding is not entirely surprising. As shown in the first section, material deprivation extends beyond income poverty: in many countries, a significant share of materially deprived children live in households that are not classified as income poor. Consequently, policies that focus narrowly on income‑based targeting may miss some families experiencing substantial living standard constraints.

Making income support more effective in reducing child material deprivation may therefore require considering eligibility criteria beyond income thresholds alone. Factors that shape the risk of deprivation could be taken into account, including income levels above the relative poverty line, family composition – particularly large families and single‑parent households – and, in some countries, the age of children, which can influence the risk of deprivation. Other indicators, such as housing cost overburden, lack of access to a vehicle, or absence of essential household appliances, could also help align support more closely with households’ actual living conditions. However, incorporating such criteria requires significant administrative capacity and may increase administrative costs. Eligibility rules must also remain sufficiently simple and transparent to avoid discouraging take‑up among the families most in need (Mroczka and Pacifico, 2025[130]) (Box 3.10).

Box 3.10. Challenges and risks associated with stronger income targeting of family benefits

Copy link to Box 3.10. Challenges and risks associated with stronger income targeting of family benefits

Targeting income support more narrowly to low-income families is often considered a way to reduce fiscal pressures while maintaining support for those most in need. Means-tested benefits can indeed concentrate resources on poorer households and improve the progressivity of social spending. In OECD countries, means-testing typically relies on a combination of household income, assets, and family composition to determine eligibility and benefit amounts. Income tests usually include labour earnings as well as other income sources such as pensions or investment income, and benefits are often gradually withdrawn as earnings rise through taper rates. Some programmes also include asset tests (Mroczka and Pacifico, 2025[130]), which consider savings or other forms of wealth, while eligibility thresholds are frequently adjusted for household size and the number of children. In certain cases, means tests interact with housing costs, employment status, or behavioural requirements such as job-search obligations, participation in training programmes, children’s school attendance, or health check-ups. While these criteria aim to improve the accuracy of targeting, they also introduce important policy and implementation challenges.

A first challenge concerns administrative complexity and implementation costs. Means-tested programmes require detailed and regularly updated information on income, assets and household composition, which increases administrative burdens for both governments and recipients. Complex eligibility rules and different withdrawal schedules across programmes can also create overlapping benefit tapers and sharp changes in disposable income as earnings rise. For instance, asset tests can increase administrative complexity and may discourage savings among low-income households, which is why some countries have relaxed or removed them in certain programmes (Joyce and Singh, 2025[131]).

Research on the Australian system, which is among the most targeted in the OECD, shows that stronger targeting assistance to the poor households can be achieved in two ways – first through income and means tests so that payments to better-off households are minimised, and second through very low levels of direct taxes on poor households, so that very little of the assistance directed to lower-income groups is clawed back (Whiteford, 2010[132]). Income support targeting also requires careful co‑ordination of withdrawal rates to avoid excessive effective marginal tax rates on additional earnings (Whiteford and Adema, 2007[65]). asset tests can increase administrative complexity and may discourage savings among low-income households, which is why some countries have relaxed or removed them in certain programmes.

A second issue relates to work incentives and potential “poverty traps.” When benefits are withdrawn as income increases, families may face high effective marginal tax rates that reduce the financial gains from additional work. While well-designed taper rates can mitigate this effect, the interaction between taxes and multiple benefits can still weaken labour supply incentives for some households (Immervoll and Pearson, 2009[133]).

Third, means testing can reduce benefit take‑up, particularly among vulnerable families. Application procedures for targeted benefits are often more complex than those for universal programmes and may require detailed financial reporting. Evidence from several OECD countries shows that a significant share of eligible households does not claim means-tested benefits because of administrative barriers, lack of information, or perceived stigma (Marc et al., 2022[134]; Ko et al., 2022[135]). Lower take‑up rates reduce the effectiveness of targeted programmes in alleviating poverty among children.

Fourth, targeted systems may face political and social sustainability challenges. Universal benefits tend to attract broader public support because they are perceived as entitlements linked to citizenship rather than assistance for a narrowly defined group. By contrast, highly targeted benefits risk losing political backing from middle‑income groups and may become more vulnerable to retrenchment over time (Korpi and Palme, 1998). This can create instability in benefit design and reduce the long-term effectiveness of child poverty strategies.

Finally, experiences with large‑scale reforms illustrate the implementation risks associated with complex targeted systems. In the United Kingdom, the introduction of the integrated means-tested benefit Universal Credit aimed to simplify the welfare system and strengthen work incentives. While the reform streamlined several programmes into a single benefit, its rollout also highlighted operational challenges, including administrative complexity, delays in payment and difficulties for some households in navigating the new system (Brewer, Joyce, Waters and Woods, 2019). Such experiences illustrate the importance of careful policy design, strong administrative capacity and gradual implementation when introducing more targeted support.

Overall, the international evidence suggests that while income targeting can help contain spending and concentrate support on families with the greatest needs, very narrow targeting can also reduce take‑up, weaken work incentives, increase administrative complexity and undermine political support for family benefits. Many OECD countries therefore rely on a mix of universal and targeted elements in family policy systems in order to balance fiscal sustainability, poverty reduction and broad social legitimacy.

The literature also points to the interplay between the volume of transfers and the degree of targeting of child benefit systems, which jointly determine the size of benefits received by families experiencing poverty or material deprivation. In systems where eligibility for child benefits is restricted to low-income households, stronger targeting is consistently associated with greater reductions in child income poverty across European countries (Van Lancker and Van Mechelen, 2015[136]). However, such “selective” systems typically operate with relatively modest redistributive budgets and therefore tend to achieve more limited overall poverty reduction. In these contexts, concentrating scarce resources on the poorest households may be the most feasible way to achieve measurable impacts.

By contrast, in universal systems the relationship between targeting and poverty reduction appears weaker and less consistent. Instead, the overall size of the redistributive budget emerges as the main driver of outcomes: larger budgets are strongly and consistently associated with greater reductions in child poverty, and universal child benefit systems typically mobilise the highest levels of spending. Some degree of targeting within universal systems (i.e. with increased amounts of child benefit for lower income families) is nonetheless associated with more generous benefit levels for low-income families, which is itself linked to stronger income poverty reduction.

Microsimulation analyses further suggest that there is no single optimal balance between universal and means-tested benefits: the most effective mix varies across countries depending on family structures, the depth of poverty, and how child poverty risks are distributed across households (González-Gallardo et al., 2024[137]).

Redesigning existing tax breaks for families with children so that they primarily benefit those experiencing income poverty and/or child material deprivation is another way to strengthen the effectiveness of income support in reducing multiple forms of child poverty. This aligns with the chapter’s findings, which show that current child-related tax breaks are only weakly associated with reductions in child poverty and may even increase spending pressures in high-spending contexts. In many countries, these measures are not designed as targeted anti-poverty instruments but take the form of non-refundable child tax credits aimed at supporting families more broadly with the costs of raising children and reducing living standard gaps between households with and without children. As a result, they tend to be ineffective in reducing child poverty, not least because low-income families often have little or no tax liability and therefore cannot fully benefit from them.

In conclusion, potential reforms should be assessed not only in terms of their impact on child income poverty but also on material deprivation, as the two are not perfectly aligned. For example, in Spain, the non-refundable child tax credit (“mínimo por descendientes”) provides greater support to higher-income households, leaving around 60% of children in poverty without access to this relief (Cruces et al., 2026[138]). Two budget-neutral reform options could be considered. One is to introduce a universal (taxable) child benefit, which has stronger potential to reduce income poverty because its taxable nature redistributes more towards lower-income families. Another is to make the existing child tax credit refundable, allowing low-income families with insufficient tax liability to receive the credit as a cash transfer. In practice, this would make it function similarly to a non-taxable, non-means-tested benefit, requiring only the introduction of refundability (as already exists for maternity and large‑family tax credits). While a universal child benefit may have greater potential to reduce income poverty (Cruces et al., 2026[138]), assessing its impact on material deprivation would provide a more comprehensive view of the relative merits of these two options.

Balancing cash transfers for families and in-kind support to improve policy effectiveness

The need to carefully balance unconditional cash transfers to families with other forms of support is suggested in this chapter’s regression analysis by the finding that a greater policy orientation towards cash benefits for families is more consistently associated with reduced child income poverty than lower severe material deprivation among children. This may reflect the fact that additional resources directed to cash benefits are more adequate in size and coverage for families below the poverty line, and that these payments could come at the expense of other types of support that are particularly effective in addressing material hardship (see Section 3.4).

Consistent with this interpretation, previous research finds that spending on sickness and disability benefits and in-kind social transfers reduces children’s risk of material deprivation more consistently than family cash benefits (Pérez-Corral and Moreno-Mínguez, 2025[139]; Guio et al., 2022[121]). Public spending on education can also mitigate deprivation by lowering households’ education-related costs (Guio et al., 2022[121]). These findings suggest that while cash transfers play an important role in supporting families’ purchasing power, they may not be sufficient if essential goods and services needed to meet children’s material needs are not available or affordable. Effectively reducing material deprivation therefore requires a balanced policy mix that combines income support with in-kind and service‑based support to ensure that the key goods and services underpinning children’s well-being are accessible.

One option to balance cash transfers and service‑based assistance are near-cash supports such as food vouchers or childcare subsidies, which are also associated with reductions in material hardship. There is suggestive evidence that targeted cash transfers (e.g. for rent or utilities) may be more effective than in-kind or provider-directed payments, as they allow households to reallocate resources according to their most pressing needs. For example, evaluations of thwinter Energy Payment in Aotearoa New Zealand found no significant reduction in reported difficulties paying utility bills among working-age beneficiaries, but did observe improvements in other dimensions of hardship, such as the ability to meet everyday expenses (Hyslop, Riggs and Maré, 2022[140]). This points to important spillover effects and highlights the value of flexibility in programme design.

Conditioning benefits payment may also strengthen the effectiveness of investments in children. For instance, linking benefit receipt to actions that improve families’ living conditions – such as housing improvements, children’s access to nutritious food programmes, regular health checks, participation in early childhood education and care (ECEC), or school attendance – can help ensure that income support contributes directly to reducing material deprivation. In this context, growing attention has been given to “cash-plus” approaches, which combine income transfers with complementary services such as childcare support, parenting programmes, employment assistance or educational support for children. These programmes aim to address multiple drivers of child deprivation simultaneously, including low household income, limited access to services and parental labour-market constraints. Evidence from experimental studies suggests that combining cash transfers with complementary services is more likely to generate sustained improvements in children’s well-being than providing cash or services alone, as it strengthens both short-term living conditions and longer-term opportunities (Bastagli et al., 2016[141]; Page, 2024[16]).

Improving take‑up of family and social benefits: evidence and policy options

Improving take‑up among disadvantaged groups remains a major challenge. The take‑up rate refers to the share of eligible families who actually claim and receive a benefit. Despite being entitled to support, many families do not take up all the benefits available to them – a phenomenon known as non-take‑up – which substantially weakens the poverty-reducing impact of social policies (Marc et al., 2022[134]). Lower socio‑economic households are less likely to receive the benefits to which they are entitled – including child and family benefits – limiting the effectiveness of transfers in reaching those most in need (OECD, 2024[142]). To address this, many OECD countries are investing in data-driven approaches that reduce administrative and informational barriers, and support proactive outreach and personalised assistance.

In response, governments and social policymakers have increasingly focussed on interventions that lower informational and administrative barriers to claiming social benefits (OECD, 2024[142]). A robust finding from experimental and evaluation research is that simplifying application processes – by reducing paperwork, streamlining eligibility criteria, and clarifying forms – lowers both information costs and procedural burdens, leading to significant increases in take‑up. A meta‑analysis of 45 randomised controlled trials across multiple countries and benefit programmes finds that behavioural interventions increase take‑up by an average of 8.1 p.p., with the largest effects observed where baseline enrolment is low. Among these interventions, simplification and personalisation are the most effective, particularly when delivered close to the moment of decision making (Daigneault et al., 2025[143]).

Automation of benefit delivery can yield even larger gains. Integrating benefit claims with existing tax or employment records to enable automatic enrolment removes the need for active application and substantially raises participation. In Malta, for example, automating access to a work-related benefit nearly quadrupled enrolment, illustrating the powerful effects of reducing administrative friction (ISSA, 2023[144]).

Uncertainty about benefit entitlements and payment adjustments can also discourage take‑up. Concerns about overpayments or future clawbacks are particularly salient for low-income and liquidity-constrained households. In the United States, for instance, few recipients of the Earned Income Tax Credit opt for early payment options, partly because of fears – sometimes unfounded – of having to repay overpaid amounts (Nichols et al., 2015[145]). Similarly, back payments may arise when benefits are insufficiently responsive to changes in household income, especially in systems without automatic adjustment mechanisms (Eurofound, 2024[146]).

Beyond simplification and automation, proactive outreach and personalised support can further improve take‑up. Field experiments show that directly contacting potentially eligible households – through targeted letters, flyers, or co‑ordinated mailings – substantially increases application rates compared with purely passive systems. In a large‑scale Belgian intervention, proactive outreach led to a three‑ to four‑fold increase in take‑up among contacted individuals (Van Gestel et al., 2023[147]). Complementary measures, such as face‑to-face assistance or guided support from caseworkers, also prove effective by helping beneficiaries understand their entitlements and navigate application procedures; experimental evidence from France shows that such support increases take‑up of income support and family benefits (Castell et al., 2025[148]). Finally, reducing geographic and procedural access barriers – by improving the local availability of services and shortening travel or processing times – can further raise participation by lowering the non-monetary costs of claiming benefits (Almeida et al., 2024[149]).

Many OECD governments are increasingly leveraging advances in technology and data to improve the coverage and delivery of social protection for vulnerable groups (OECD, 2024[142]; 2025[18]). These efforts include using probabilistic estimates based on (often de‑identified) survey and administrative data to better identify disadvantaged groups and regions in need, tailor communication strategies, and guide investments in new or expanded programmes. Countries are also making greater use of linked administrative data – often through unique personal identifiers or social registries – to assess benefit eligibility at the individual level. This can help identify gaps in coverage (for example, by proactively encouraging eligible households to apply) and reduce administrative burdens on beneficiaries, such as through pre‑filled applications drawing on existing records.

At the same time, the use of technology and data raises important challenges and risks. Effective implementation requires strong cross-government co‑ordination, robust data governance, and strict safeguards to protect individuals’ privacy. Governments must also carefully manage the risk that automated systems and data-driven decision making embed or amplify discriminatory biases, which could reinforce existing inequalities or create new forms of exclusion and disadvantage (OECD, 2024[142]).

Designing benefit payment structure to enhance poverty reduction

Predictable and regular cash transfers, such as monthly or annual payments, play a crucial role in enabling families to plan and meet their needs. Regular and stable income flows that families can reasonably expect to persist over time enable families to plan expenditures, manage resources, and invest in their children’s development, and thus constitute a key mechanism through which income influences child outcomes. The frequency and structure of payments matter: regular monthly transfers are more effective at easing ongoing financial pressures – such as food insecurity or difficulties paying rent –, which are especially pronounced among low-income households and those facing acute economic disadvantage. By contrast, less frequent but larger payments are more likely to finance substantial, often postponed expenditures, such as medical care, home improvements or durable goods, thereby helping households address forms of hardship associated with delayed essential consumption. However, evidence suggests that regular and predictable income transfers sustained throughout childhood have stronger and more lasting effects. By providing stable financial resources over time, they are more effective at consistently easing chronic financial stress and enabling parents to make sustained investments in their children’s development, with benefits that extend into adulthood through greater expected lifetime earnings (Page, 2024[16]).

In addition, the timeliness of payments matters. Timely support is essential to ensure that assistance reaches families when it is most needed – ideally as close as possible to the moment they experience an income shock or other adverse event that constrains their ability to meet immediate expenses, such as paying bills or purchasing essential goods and services required for children’s well-being. This is particularly important given that, in many countries, the share of “asset-poor” individuals – those lacking sufficient liquid financial assets to maintain a living standard above the poverty threshold for at least three months – significantly exceeds the share of those who are income‑poor (Balestra and Oehler, 2023[150]).

Fragmentation and complexity in benefit systems can reduce transparency and make support difficult to navigate. Strengthening co‑ordination – for example, through a universal base benefit complemented by predictable, needs-based top-ups – can enhance both clarity and stability. Such an approach helps ensure that support is reliable and aligned with families’ circumstances, thereby improving its effectiveness in supporting children’s well-being.

Designing income support involves balancing trade‑offs across three key parameters: the assessment period, the payment frequency, and the duration for which benefits are fixed before reassessment (Millar et al., 2020[151]). Short assessment periods can be unrepresentative of typical earnings and may be open to manipulation, while longer periods better smooth income but may be less responsive to recent changes. Similarly, frequent payments provide more regular and stable income, whereas less frequent payments can make budgeting harder. Short award durations increase responsiveness but can lead to volatile payments, while longer durations offer stability but risk misalignment between benefits and current needs.

In practice, overly frequent reassessments of means-tested benefits can undermine this stability, creating volatility and financial stress – especially for families with fluctuating earnings. For example, a recent report on child benefits in Norway highlights that housing support is assessed using monthly income information, leading to substantial month-to-month variation in payments for families with volatile earnings (Kunnskapsdepartementet, 2024[152]). To improve predictability, the expert group recommends maintaining housing support for families with children over a fixed six‑month period, even when income fluctuates, thereby strengthening income stability and supporting better financial planning.

More generally, a central trade‑off emerges between responsiveness and income security. While adjusting benefits quickly to changing circumstances improves targeting, it can also create financial instability, stress, and administrative burden. Given that income and household situations for low-income families often fluctuate temporarily rather than permanently, frequent reassessments can lead to “churning” in and out of eligibility without reflecting fundamental changes in need. Balancing accuracy, simplicity, and stability is therefore key in benefit design.

As highlighted by Millar and Whiteford (2020[151]), recent reforms in Australia and the United Kingdom illustrate how design choices can inadvertently generate overpayments. In Australia, reliance on estimated annual income often led to underestimation and thus overpayments, while in the United Kingdom, retrospective income assessments failed to capture changes during the year. Advances in data matching and automation have made it easier to detect and recover overpayments, sometimes leading to problematic practices such as large‑scale debt recovery.

Governments can mitigate their impact through policy choices such as income disregards, waivers, lump-sum adjustments, or extended repayment periods. By contrast, some design features – particularly in the UK’s Universal Credit system – have directly increased the risk of debt. Monthly payments in arrears, initial waiting periods, and the use of repayable advance payments mean many claimants must rely on borrowing. Other features, such as paying housing support directly to claimants, can further raise the risk of arrears. Overall, while accurate reconciliation is important, many overpayments and resulting debts stem from deliberate design choices, often placing a disproportionate burden on low-income households (Millar et al., 2020[151]). Preventing such issues through better system design is preferable to relying on ex post corrections.

Comprehensive policy action to reduce child poverty also requires effective local services

Complementing individual income support and service subsidies discussed in this report with strong place‑based policies is critical to addressing geographical inequalities in access to essential services for children and caregivers (OECD, 2025[153]; 2025[13]). Income support alone is unlikely to be fully effective – or even fully taken up – if adequate services are not available in the areas where disadvantaged families live. Place‑based approaches are particularly important for addressing shortages in ECEC and healthcare provision in deprived regions and neighbourhoods, improving poor-quality housing, and reducing forms of material deprivation linked to inadequate local services, spatial segregation, or weak transport infrastructure limiting access to childcare, healthcare, leisure, sport, and play facilities and green spaces that are essential for children’s and caregivers’ well-being (OECD, 2025[153]). This is one reason why place‑based policies are highlighted as an avenue to strengthening the implementation of the European Child Guarantee (European Commission, 2026[154]).

At the same time, implementing effective place‑based policies raises important governance and capacity challenges. It requires identifying local needs accurately, assessing whether the local supply of services is adequate in terms of both affordability and geographical accessibility, and building the capacity of local authorities to monitor these dimensions over time. Achieving this also depends on developing shared frameworks and common standards across territories to assess needs and service provision, enabling both local and national authorities to set priorities, allocate resources effectively, and monitor progress consistently (OECD, 2025[155]; 2025[18]).

References

[66] Adema, W., C. Clarke and O. Thévenon (2020), “Family Policies and Family Outcomes in OECD Countries”, in Nieuwenhuis, R. and W. Van Lancker (eds.), The Palgrave Handbook of Family Policy, Palgrave Macmillan, https://doi.org/10.1007/978-3-030-54618-2_9.

[117] Aizer, A. et al. (2016), “The Long-Run Impact of Cash Transfers to Poor Families”, American Economic Review, Vol. 106/4, pp. 935-71, https://doi.org/10.1257/AER.20140529.

[149] Almeida, V. et al. (2024), “Geographic inequalities in accessibility of essential services”, OECD Social, Employment and Migration Working Papers, No. 307, OECD Publishing, Paris, https://doi.org/10.1787/12bab9fb-en.

[79] Ananat, E. et al. (2024), “Effects of the expanded Child Tax Credit on employment outcomes”, Journal of Public Economics, Vol. 238, p. 105168, https://doi.org/10.1016/J.JPUBECO.2024.105168.

[95] Atkinson, A. (2011), “The case for universal child benefit”, in Walker, A., A. Sinfield and C. Walker (eds.), Fighting Poverty, Inequality and Injustice: A Manifesto Inspired by Peter Townsend, Bristol University Press, https://doi.org/10.46692/9781847427168.005.

[88] Ayala, L. and M. Rodríguez (2013), “Evaluating social assistance reforms under programme heterogeneity and alternative measures of success”, International Journal of Social Welfare, Vol. 22/4, pp. 406-419, https://doi.org/10.1111/IJSW.12008.

[64] Bäckman, O. and T. Ferrarini (2010), “Combating Child Poverty? A Multilevel Assessment of Family Policy Institutions and Child Poverty in 21 Old and New Welfare States”, Journal of Social Policy, Vol. 39/02, p. 275, https://doi.org/10.1017/S0047279409990456.

[86] Balestra, C., D. Hirsch and D. Vaughan-Whitehead (2023), “Living wages in context: A comparative analysis for OECD countries”, OECD Papers on Well-being and Inequalities, No. 13, OECD Publishing, Paris, https://doi.org/10.1787/2e622174-en.

[150] Balestra, C. and F. Oehler (2023), “Measuring the joint distribution of household income, consumption and wealth at the micro-level”, OECD Papers on Well-being and Inequalities, No. 11, OECD Publishing, Paris, https://doi.org/10.1787/f9d85db6-en.

[85] Bárcena-Martín, E., M. Blanco-Arana and S. Pérez-Moreno (2018), “Social Transfers and Child Poverty in European Countries: Pro-poor Targeting or Pro-child Targeting?”, Journal of Social Policy, Vol. 47/4, pp. 739-758, https://doi.org/10.1017/S0047279418000090.

[92] Bárcena-Martín, E. et al. (2017), “Child deprivation and social benefits: Europe in cross-national perspective”, Socio-Economic Review, Vol. 15/4, pp. 717-744, https://doi.org/10.1093/SER/MWX019.

[102] Barrow, L. and L. McGranahan (2000), “The Effects of the Earned Income Credit on the Seasonality of Household Expenditures”, National Tax Journal, Vol. 53/4, pp. 1211-1244, https://doi.org/10.17310/NTJ.2000.4S1.08.

[141] Bastagli, F. et al. (2016), “Cash transfers: what does the evidence say? A rigorous review of programme impact and of the role of design and implementation features”, https://odi.org/en/publications/cash-transfers-what-does-the-evidence-say-a-rigorous-review-of-impacts-and-the-role-of-design-and-implementation-features/ (accessed on 22 January 2021).

[41] Bouget, D. et al. (2015), Social investment in Europe – A study of national policies 2015, European Commission: Directorate-General for Employment, Social Affairs and Inclusion, Liser, European Social Observatory, Applica, Publications Office, https://data.europa.eu/doi/10.2767/084978 (accessed on 12 September 2025).

[57] Card, D., J. Kluve and A. Weber (2017), “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations”, Journal of the European Economic Association, Vol. 16/3, pp. 894-931, https://doi.org/10.1093/jeea/jvx028.

[148] Castell, L. et al. (2025), “Take-up of Social Benefits: Experimental Evidence from France”, American Economic Journal: Economic Policy, Vol. 17/4, pp. 1-29, https://doi.org/10.1257/POL.20220786.

[12] Chetty, R. and N. Hendren (2018), “The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects*”, The Quarterly Journal of Economics, Vol. 133/3, pp. 1107-1162, https://doi.org/10.1093/qje/qjy007.

[11] Chetty, R. et al. (2014), “Where is the land of Opportunity? The Geography of Intergenerational Mobility in the United States *”, The Quarterly Journal of Economics, Vol. 129/4, pp. 1553-1623, https://doi.org/10.1093/qje/qju022.

[96] Chzhen, Y. (2017), “Unemployment, social protection spending and child poverty in the European Union during the Great Recession”, Journal of European Social Policy, Vol. 27/2, pp. 123-137, https://doi.org/10.1177/0958928716676549.

[3] Clarke, C. et al. (2022), “The economic costs of childhood socio-economic disadvantage in European OECD countries”, OECD Papers on Well-being and Inequalities, No. 9, OECD Publishing, Paris, https://doi.org/10.1787/8c0c66b9-en.

[169] Cleiss (2026), The French Social Security System - IV - Family benefits, Centre des liaisons européennes et internationales de sécurité sociale (Cleiss), https://www.cleiss.fr/docs/regimes/regime_france/an_4.html (accessed on 14 September 2026).

[111] Collyer Sophie, C. Wimer and D. Harris (2022), Keeping Up with Inflation: How Policy Indexation Can Enhance Poverty Reduction, The Century Foundation, https://tcf.org/content/report/keeping-up-with-inflation-how-policy-indexation-can-enhance-poverty-reduction/ (accessed on 15 September 2026).

[10] Cooper, K. and K. Stewart (2021), “Does Household Income Affect children’s Outcomes? A Systematic Review of the Evidence”, Child Indicators Research, Vol. 14/3, pp. 981-1005, https://doi.org/10.1007/S12187-020-09782-0/.

[15] Cooper, K. and K. Stewart (2017), “Does Money Affect Children’s Outcomes? An update”, CASEpapers 203, Centre for Analysis of Social Exclusion, https://researchonline.lse.ac.uk/id/eprint/103494.

[4] Corak, M. (2013), “Income Inequality, Equality of Opportunity, and Intergenerational Mobility”, Journal of Economic Perspectives, Vol. 27/3, pp. 79-102, https://doi.org/10.1257/JEP.27.3.79.

[84] Corluy, V. and F. Vandenbroucke (2017), “Individual Employment, Household Employment and Risk of Poverty in the EU. A Decomposition Analysis”, in Atkinson, A., A. Guio and E. Marlier (eds.), Monitoring Social Inclusion in Europe, Publications Office of the European Union , Luxembourg, https://papers.ssrn.com/abstract=2981712. (accessed on 17 September 2025).

[138] Cruces, H. et al. (2026), Spain’s Child Support Fails the Poor: Actionable Policy Reforms, European Commission, Seville, https://publications.jrc.ec.europa.eu/repository/handle/JRC146328.

[143] Daigneault, P. et al. (2025), “How effective are behavioral interventions to increase the take-up of social benefits? A systematic review of field experiments”, Journal of Policy Analysis and Management, Vol. 44/3, pp. 997-1059, https://doi.org/10.1002/PAM.70005.

[21] Daly, M. and S. Ryu (2025), “Child-related Social Policies in Europe during the COVID-19 Pandemic”, Journal of Social Policy, Vol. 54/3, pp. 894-911, https://doi.org/10.1017/S0047279423000351.

[89] Danziger, S. et al. (2000), “Work, income, and material hardship after welfare reform”, Journal of Consumer Affairs, Vol. 34/1, pp. 6-30, https://doi.org/10.1111/J.1745-6606.2000.TB00081.X.

[90] Danziger, S. et al. (2002), “Does It Pay to Move from Welfare to Work?”, Journal of Policy Analysis and Management, Vol. 21/4, pp. 671-692, https://doi.org/10.1002/PAM.10080.

[101] Davison, G. et al. (2018), “The State of State EITCs: An Overview and Their Implications for Low- and Moderate-Income Households”, Center for Social Development Research, https://doi.org/10.7936/K7WM1CZ0.

[6] De Schutter, O. et al. (2023), The Escape from Poverty: Breaking the Vicious Cycles Perpetuating Disadvantage, Policy Press, Bristol, https://doi.org/10.51952/9781447370611.

[20] Dely, B., R. Hyee and C. Prinz (2025), “What worked well in social protection during the COVID-19 pandemic?”, OECD Social, Employment and Migration Working Papers, No. 325, OECD Publishing, Paris, https://doi.org/10.1787/b31ac155-en.

[59] Diris, R., F. Vandenbroucke and G. Verbist (2017), “The impact of pensions, transfers and taxes on child poverty in Europe: the role of size, pro-poorness and child orientation”, Socio-Economic Review, Vol. 15/4, p. mww045, https://doi.org/10.1093/ser/mww045.

[109] Duncan, G., P. Morris and C. Rodrigues (2011), “Does Money Really Matter? Estimating Impacts of Family Income on Young Children’s Achievement With Data From Random-Assignment Experiments”, Developmental Psychology, Vol. 47/5, pp. 1263-1279, https://doi.org/10.1037/A0023875.

[1] Duncan, G. et al. (1998), “How much does childhood poverty affect the life chances of children?”, American sociological review, pp. 406-423, https://doi.org/10.2307/2657556.

[56] Égert, B., J. Botev and D. Turner (2020), “The contribution of human capital and its policies to per capita income in Europe and the OECD”, European Economic Review, Vol. 129, p. 103560, https://doi.org/10.1016/j.euroecorev.2020.103560.

[39] Esping-Andersen, Gosta (2002), “Why We Need a New Welfare State”, Why We Need a New Welfare State, https://doi.org/10.1093/0199256438.001.0001.

[146] Eurofound (2024), Social protection 2.0: Unemployment and minimum income benefits, Publications Office of the European Union, Luxembourg, https://doi.org/10.2806/0704651.

[154] European Commission (2026), Breaking the cycle of child poverty: strengthening the European Child Guarantee, DG Employment, Social Affairs and Inclusion - Staff working document.

[54] European Commission (2026), Commission proposes new measures to fight poverty and improve lives of persons with disabilities, European Commission, Directorate-General for Employment, Social Affairs and Inclusion, https://employment-social-affairs.ec.europa.eu/news/commission-proposes-new-measures-fight-poverty-and-improve-lives-persons-disabilities-2026-05-06_en (accessed on 11 September 2026).

[51] European Commission (2026), Communication on a strengthened European Child Guarantee, https://employment-social-affairs.ec.europa.eu/document/7f60af79-936f-42b3-a6c4-a4639c6679ff_en (accessed on 14 September 2026).

[50] European Commission (2026), Proposal for a Council recommendation on fighting housing exclusion, https://employment-social-affairs.ec.europa.eu/document/21450169-5eaf-40e7-8a5a-4133f20510f1_en (accessed on 14 September 2026).

[42] European Commission (2024), Employment and social developments in Europe 2024, Directorate-General for Employment, Social Affairs and Inclusion. Publications Office of the European Union, https://data.europa.eu/doi/10.2767/91555 (accessed on 12 September 2025).

[48] European Commission (2021), European Child Guarantee, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32021H1004&from=EN.

[45] European Commission (2013), Recommendation on Investing in children: breaking the cycle of disadvantage, https://employment-social-affairs.ec.europa.eu/policies-and-activities/social-protection-social-inclusion/addressing-poverty-and-supporting-social-inclusion/investing-children_en.

[44] European Commission (n.d.), Addressing poverty and supporting social inclusion, https://employment-social-affairs.ec.europa.eu/policies-and-activities/social-protection-social-inclusion/addressing-poverty-and-supporting-social-inclusion_en (accessed on 14 September 2026).

[53] European Commission (n.d.), Communication “Enhancing the strategy for the rights of persons with disabilities up to 2030“, https://commission.europa.eu/document/ab0bcea9-6b3a-48c1-875c-1cc687fba6d0_en.

[49] European Commission (n.d.), EU Anti-Poverty Strategy, https://employment-social-affairs.ec.europa.eu/policies-and-activities/social-protection-social-inclusion/addressing-poverty-and-supporting-social-inclusion/eu-anti-poverty-strategy_en (accessed on 14 September 2026).

[52] European Commission (n.d.), European Child Guarantee, https://employment-social-affairs.ec.europa.eu/policies-and-activities/social-protection-social-inclusion/addressing-poverty-and-supporting-social-inclusion/investing-children/european-child-guarantee_en (accessed on 14 September 2026).

[46] European Commission (n.d.), European Pillar of Social Rights, https://employment-social-affairs.ec.europa.eu/policies-and-activities/european-pillar-social-rights-building-fairer-and-more-inclusive-european-union_en (accessed on 14 September 2026).

[47] European Commission (n.d.), European Pillar of Social Rights Action Plan, https://employment-social-affairs.ec.europa.eu/policies-and-activities/european-pillar-social-rights-building-fairer-and-more-inclusive-european-union/european-pillar-social-rights-action-plan_en (accessed on 14 September 2026).

[98] European Commission: Directorate-General for Employment, S. (2026), Report on the effectiveness of child benefits in the EU – Final report, Publications Office of the European Union, Paris, https://data.europa.eu/doi/10.2767/1749160 (accessed on 16 September 2026).

[156] Eurostat (2026), Glossary: Material deprivation, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Glossary:Material_deprivation.

[27] Eurostat (2023), Investing in Children, https://ec.europa.eu/eurostat/web/employment-and-social-inclusion-indicators/social-protection-and-inclusion/investing-children.

[25] Eurostat (n.d.), European Union Statistics on Income and Living Conditions (EU-SILC) survey, https://ec.europa.eu/eurostat/web/microdata/collections-research/european-union-statistics-on-income-and-living-conditions.

[28] Eurostat (n.d.), Glossary: Child deprivation, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Glossary:Child_deprivation (accessed on 11 September 2026).

[157] Eurostat (n.d.), Population and social conditions, https://ec.europa.eu/eurostat/databrowser/product/page/ILC_MDDD11 (accessed on 14 September 2026).

[173] Eurostat (n.d.), Statistics Explained - Your guide to European statistics, https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Main_Page (accessed on 14 September 2026).

[104] Gao, Q., N. Kaushal and J. Waldfogel (2009), “How have expansions in the EITC affected family expenditures?”, in Ziliak, J. (ed.), Welfare Reform and Its Long-Term Consequences for America’s Poor, Cambridge University Press, Cambridge, UK.

[32] Global coalition to end child poverty (2025), Countries commitment to end child poverty: A review of the Voluntary National Reviews submitted from 2017 to 2024, Global coalition to end child poverty, https://www.endchildhoodpoverty.org/publications-feed/vnranalysis (accessed on 12 September 2025).

[78] Goldin, J. et al. (2024), “Work Requirements and Child Tax Benefits”, https://doi.org/10.3386/W32343.

[168] Gonzales, F. (2025), “Canada adjusts child benefit as inflation indexation takes effect”, Benefits and Pensions Monitor, https://www.benefitsandpensionsmonitor.com/news/industry-news/canada-adjusts-child-benefit-as-inflation-indexation-takes-effect/392251 (accessed on 14 September 2026).

[137] González-Gallardo, S. et al. (2024), “Social transfer policies and child poverty in European countries: Evaluating policy alternatives through a multiobjective programming model”, Socio-Economic Planning Sciences, Vol. 93, p. 101892, https://doi.org/10.1016/J.SEPS.2024.101892.

[103] Goodman-Bacon, A. and L. McGranahan (2008), “How do EITC recipients spend their refunds?”, Economic Perspectives, Vol. 32/Q II, pp. 17-32, https://ideas.repec.org/a/fip/fedhep/y2008iqiip17-32nv.32no.2.html (accessed on 15 September 2026).

[35] Government of Canada (2018), Opportunity for All - Canada’s First Poverty Reduction Strategy, Employment and Social Development Canada, https://www.canada.ca/en/employment-social-development/programs/poverty-reduction/reports/strategy.html (accessed on 24 March 2026).

[37] Government of New Zealand (2025), Putting the Strategy into action - Taking a social investment approach, Ministry of Social Development, https://www.msd.govt.nz/about-msd-and-our-work/child-wellbeing-and-poverty-reduction/child-youth-strategy/putting-the-strategy-into-action.html (accessed on 12 September 2025).

[36] Government of New Zealand (2024), The Child and Youth Strategy 2024-27, Ministry of Social Development, https://www.msd.govt.nz/about-msd-and-our-work/child-wellbeing-and-poverty-reduction/child-youth-strategy/index.html.

[91] Grant, M. and K. Prickett (2023), Research evidence on policy supports and the effect on material hardship, Department of Prime Minister and Cabinet, Wellington, New Zealand.

[26] Guio, A. et al. (2018), “Towards an EU measure of child deprivation”, Child Indicators Research, Vol. 11/3, pp. 835-860, https://doi.org/10.1007/s12187-017-9491-6.

[121] Guio, A. et al. (2022), “Differences in Child Deprivation Across Europe: The Role of In-Cash and In-Kind Transfers”, Child Indicators Research 2022 15:6, Vol. 15/6, pp. 2363-2388, https://doi.org/10.1007/s12187-022-09948-y.

[119] Guldi, M. et al. (2022), “Supplemental Security Income for Children, Maternal Labor Supply, and Family Well-Being: Evidence from Birth Weight Eligibility Cutoffs”, Journal of Human Resources, Vol. 59/4, pp. 975-1010, https://doi.org/10.3368/JHR.0818-9654R2.

[114] Han, J. (2016), “The impact of SNAP on material hardships: Evidence from broad-based categorical eligibility expansions”, Southern Economic Journal, Vol. 83/2, pp. 464-486, https://doi.org/10.1002/SOEJ.12171;WGROUP:STRING:PUBLICATION.

[127] HCFEA (2023), Le pouvoir d’achat des familles face au choc d’inflation, Haut Conseil à la Famille, l’Enfance and l’Âge, https://www.strategie-plan.gouv.fr/publications/hcfea-pouvoir-dachat-familles-face-choc-dinflation (accessed on 6 January 2026).

[67] Hemerijck, A., S. Ronchi and I. Plavgo (2023), “Social investment as a conceptual framework for analysing well-being returns and reforms in 21st century welfare states”, Socio-Economic Review, Vol. 21/1, pp. 479-500, https://doi.org/10.1093/SER/MWAC035.

[23] Hiilamo, A. and T. Lallukka (2023), “Child poverty is exacerbated by the current war-induced cost-of-living crisis: need for new European-wide measures”, European Journal of Public Health, Vol. 33/3, pp. 353-353, https://doi.org/10.1093/EURPUB/CKAD021.

[38] HM Government (2025), Our Children, Our Future: Tackling Child Poverty, Department of Work and Pension, London, http://www.gov.uk/official-documents (accessed on 19 February 2026).

[61] Hobcraft, J. and K. Kiernan (2001), “Childhood poverty, early motherhood and adult social exclusion”, The British Journal of Sociology, Vol. 52/3, pp. 495-517, https://doi.org/10.1080/00071310120071151.

[107] Hoynes, H. (2019), “The Earned Income Tax Credit”, Annals of the American Academy of Political and Social Science, Vol. 686/1, pp. 180-203, https://doi.org/10.1177/0002716219881621;SUBPAGE:STRING:FULL.

[108] Hoynes, H. and J. Rothstein (2017), “Tax Policy toward Low-Income Families”, in Auerbach, A. and K. Smetters (eds.), The Economics of Tax Policy, The Economics of Tax Policy, Oxford University Press.

[115] Hoynes, H., D. Schanzenbach and D. Almond (2016), “Long-Run Impacts of Childhood Access to the Safety Net”, American Economic Review, Vol. 106/4, pp. 903-934, https://doi.org/10.1257/aer.20130375.

[140] Hyslop, D., L. Riggs and D. Maré (2022), “The impact of the 2018 Families Package Winter Energy Payment policy”, https://www.stats.govt.nz/integrated-data/. (accessed on 6 May 2026).

[128] IGAS (2021), Revue des dépenses socio-fiscales en faveur de la politique familiale, Inspection Générale des Affaires Sociales, Paris, https://www.igf.finances.gouv.fr/igf/accueil/nos-activites/rapports-de-missions/liste-de-tous-les-rapports-de-mi/revue-des-depenses-socio-fiscale.html (accessed on 6 January 2026).

[93] ILO, UNICEF and Learning for Well-Being Institute (2024), “The promise of universal child benefits: The foundational policy for economic and social development”, Policy brief, https://doi.org/10.54394/YSAT8221.

[133] Immervoll, H. and M. Pearson (2009), “A Good Time for Making Work Pay? Taking Stock of In-Work Benefits and Related Measures across the OECD”, OECD Social, Employment and Migration Working Papers, No. 81, OECD Publishing, Paris, https://doi.org/10.1787/225442803245.

[144] ISSA (2023), Sécurité sociale et droits humains, accès aux droits et lutte contre le non-recours, https://www.issa.int/fr/analysis/social-security-and-human-rights-ensuring-access-and-combatting-non-take-social-benefits (accessed on 6 January 2026).

[116] Jacob, B., M. Kapustin and J. Ludwig (2015), “The Impact of Housing Assistance on Child Outcomes: Evidence from a Randomized Housing Lottery”, The Quarterly Journal of Economics, Vol. 130/1, pp. 465-506, https://doi.org/10.1093/QJE/QJU030.

[131] Joyce, M. and A. Singh (2025), “Income uncertainty, precautionary wealth, and social insurance”, European Economic Review, Vol. 175/1, p. 105003, https://doi.org/10.1016/j.euroecorev.2025.105003.

[82] Koebel, K. and M. Stabile (2024), Child benefits in an international comparative context, VOXEU Column, https://cepr.org/voxeu/columns/child-benefits-international-comparative-context (accessed on 9 January 2026).

[129] Koebel, K. and M. Stabile (2023), “Evaluating the Effects of the 2021 Expansion of the Child Tax Credit: The International Comparative Context”, Annals of the American Academy of Political and Social Science, Vol. 710/1, pp. 36-56, https://doi.org/10.1177/0002716224125883.

[105] Kondratjeva, O. et al. (2021), “The impact of state earned income tax credit increases on material and medical hardship”, Journal of Consumer Affairs, Vol. 55/3, pp. 872-910, https://doi.org/10.1111/JOCA.12382.

[135] Ko, W. et al. (2022), “Take-up of social benefits”, NBER Working Papers Series, No. 30148, https://doi.org/10.3386/w30148.

[166] Kropko, J. et al. (2014), “Multiple Imputation for Continuous and Categorical Data: Comparing Joint Multivariate Normal and Conditional Approaches”, Political Analysis, Vol. 22/4, pp. 497-519, https://doi.org/10.1093/PAN/MPU007.

[152] Kunnskapsdepartementet (2024), Et jevnere utdanningsløp, https://www.regjeringen.no/no/dokumenter/et-jevnere-utdanningslop/id3025513/ (accessed on 9 January 2026).

[31] LIS Cross-National Data Centre (n.d.), Luxembourg Income Study, https://www.lisdatacenter.org/our-data/lis-database/.

[55] Lorenzoni, L. et al. (2018), “Which policies increase value for money in health care?”, OECD Health Working Papers, No. 104, OECD Publishing, Paris, https://doi.org/10.1787/a46c5b1f-en.

[62] Maldonado, L. and R. Nieuwenhuis (2015), “Family policies and single parent poverty in 18 OECD countries, 1978–2008”, Community, Work & Family, Vol. 18/4, pp. 395-415, https://doi.org/10.1080/13668803.2015.1080661.

[134] Marc, C. et al. (2022), Non-take-up of minimum social benefits: quantification in Europe A salient phenomenon still not making public policy headlines.

[76] Marr, C. et al. (2015), “EITC and Child Tax Credit Promote Work, Reduce Poverty, and Support Children’s Development”, Center on Budget and Policy Priorities, https://www.cbpp.org/research/federal-tax/eitc-and-child-tax-credit-promote-work-reduce-poverty-and-support-childrens (accessed on 9 January 2026).

[14] Masarik, A. and R. Conger (2017), “Stress and child development: a review of the Family Stress Model”, Current Opinion in Psychology, Vol. 13, pp. 85-90, https://doi.org/10.1016/J.COPSYC.2016.05.008.

[77] McInnis, N. et al. (2023), “The Intergenerational Transmission of Poverty and Public Assistance: Evidence from the Earned Income Tax Credit”, https://doi.org/10.3386/W31429.

[75] Michelmore, K. (2025), “Tax Credits and Child Outcomes: Lessons from the U.S., U.K., and Canada”, https://doi.org/10.3386/W33822.

[120] Miho, A. and O. Thévenon (2020), “Treating all children equally?: Why policies should adapt to evolving family living arrangements”, OECD Social, Employment and Migration Working Papers, No. 240, OECD Publishing, Paris, https://doi.org/10.1787/83307d97-en.

[151] Millar, J. et al. (2020), “Timing it right or timing it wrong: how should income-tested benefits deal with changes in circumstances?:”, Journal of Poverty and Social Justice, Vol. 28/1, pp. 3-20, https://doi.org/10.1332/175982719X15723525915871.

[106] Miller, C. et al. (2018), “Boosting The Earned Income Tax Credit For Singles: Final Impact Findings from the Paycheck Plus Demonstration in New York City”.

[100] Milligan, K. and M. Stabile (2011), “Do Child Tax Benefits Affect the Well-Being of Children? Evidence from Canadian Child Benefit Expansions”, American Economic Journal: Economic Policy, Vol. 3/3, pp. 175-205, https://doi.org/10.1257/POL.3.3.175.

[99] Moffitt, R. (2015), “The Deserving Poor, the Family, and the U.S. Welfare System”, Demography, Vol. 52/3, pp. 729-749, https://doi.org/10.1007/s13524-015-0395-0.

[164] Moritz, S. and T. Bartz-Beielstein (2017), “imputeTS: Time Series Missing Value Imputation in R”, The R Journal, Vol. 9/1, p. 207, https://doi.org/10.32614/rj-2017-009.

[130] Mroczka, J. and D. Pacifico (2025), “Comparative Assessment of Asset Requirements for Minimum Income Benefit Recipients”, OECD Social, Employment and Migration Working Papers, No. 328, OECD Publishing, Paris, https://doi.org/10.1787/73fbcd99-en.

[81] Mroczka, J., D. Pacifico and E. Chen (2025), “Comparative assessment of activation requirements for unemployment and minimum income benefit recipients”, OECD Social, Employment and Migration Working Papers, No. 329, OECD Publishing, Paris, https://doi.org/10.1787/86ec18af-en.

[2] National Academies of Sciences, Engineering and Medicine (2019), A Roadmap to Reducing Child Poverty, The National Academies Press, Washington, https://doi.org/10.17226/25246.

[145] Nichols, A. et al. (2015), “The Earned Income Tax Credit (EITC)”, http://www.nber.org/papers/w21211 (accessed on 6 January 2026).

[74] Nieuwenhuis, R. (2022), “No activation without reconciliation? The interplay between ALMP and ECEC in relation to women’s employment, unemployment and inactivity in 30 OECD countries, 1985–2018”, Social Policy and Administration, Vol. 56/5, pp. 808-826, https://doi.org/10.1111/SPOL.12806.

[87] OECD (2026), “The wide-ranging effects of active labour market policies in Slovenia: Lessons learnt and implications for other countries”, https://www.oecd.org/content/dam/oecd/en/about/projects/technical-reports-and-presentations-dg- (accessed on 22 April 2026).

[155] OECD (2025), Building brighter futures − Implementing the European Child Guarantee in Portugal, OECD Publishing, Paris, https://doi.org/10.1787/bbc928ea-en.

[153] OECD (2025), “Building child-friendly neighbourhoods: Empowering communities with data-driven action”, OECD Papers on Well-being and Inequalities, No. 33, OECD Publishing, Paris, https://doi.org/10.1787/4f46a260-en.

[18] OECD (2025), How to monitor and evaluate policies to address child poverty and disadvantage in Ireland, OECD Publishing, Paris, https://doi.org/10.1787/6f71f7bb-en.

[172] OECD (2025), Pensions at a Glance 2025: OECD and G20 Indicators, OECD Publishing, Paris, https://doi.org/10.1787/e40274c1-en.

[70] OECD (2025), Reducing Inequalities by Investing in Early Childhood Education and Care, Starting Strong, OECD Publishing, Paris, https://doi.org/10.1787/b78f8b25-en.

[30] OECD (2025), The cost of raising children across evolving family structures: New estimates and policy challenges, OECD, Paris, https://doi.org/10.1787/6550982c-en.

[63] OECD (2025), The Economic Benefit of Promoting Healthy Ageing and Community Care, OECD Health Policy Studies, OECD Publishing, Paris, https://doi.org/10.1787/0f7bc62b-en.

[13] OECD (2025), “The importance of monitoring neighbourhood conditions for children’s well-being and development”, OECD Papers on Well-being and Inequalities, No. 34, OECD Publishing, Paris, https://doi.org/10.1787/143ae959-en.

[142] OECD (2024), Modernising Access to Social Protection: Strategies, Technologies and Data Advances in OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/af31746d-en.

[122] OECD (2024), OECD Tax Ben, Comparative policy tables, https://webfs.oecd.org/Els-com/Tax Ben/Comparative policy tables/OECD-Policy-tables-2024.xlsx.

[170] OECD (2024), Tax and benefit policy descriptions for Austria 2024, https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Austria-latest.pdf.

[167] OECD (2024), Tax and benefit policy descriptions for Belgium 2024, https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Belgium-latest.pdf (accessed on 14 September 2026).

[161] OECD (2024), TaxBEN: The OECD tax-benefit simulation model. Methodology, user guide and policy applications, https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/income-support-redistribution-and-work-incentives/OECD-TaxBEN-methodology-and-manual.pdf.

[126] OECD (2023), Taxing Wages 2023: Indexation of Labour Taxation and Benefits in OECD Countries, OECD Publishing, Paris, https://doi.org/10.1787/8c99fa4d-en.

[71] OECD (2020), Is Childcare affordable?, OECD, Paris, http://oe.cd/childcare-brief-2020 (accessed on 6 January 2021).

[5] OECD (2018), A Broken Social Elevator? How to Promote Social Mobility, OECD Publishing, Paris, https://doi.org/10.1787/9789264301085-en.

[33] OECD (2018), Poor children in rich countries: why we need policy action, OECD, http://www.oecd.org/social/family/Poor-children-in-rich-countries-Policy-brief-2018.pdf.

[83] OECD (2015), “The quality of working lives: Earnings mobility, labour market risk and long-term inequality”, in OECD Employment Outlook, OECD Publishing, Paris, https://doi.org/10.1787/empl_outlook-2015-en (accessed on 16 September 2025).

[171] OECD (2014), Tax and benefit policy descriptions for Luxembourg 2024, https://www.oecd.org/content/dam/oecd/en/topics/policy-sub-issues/incomes-support-redistribution-and-work-incentives/TaxBEN-Luxembourg-latest.pdf.

[7] OECD (2011), Doing Better for Families, OECD Publishing, Paris, https://doi.org/10.1787/9789264098732-en.

[124] OECD (n.d.), How do countries calculate tax liabilities and social benefit entitlements?, https://www.oecd.org/en/topics/sub-issues/income-support-redistribution-and-work-incentives/how-do-countries-calculate-tax-liabilities-and-social-benefit-entitlements.html.

[19] OECD (n.d.), Income distribution database, https://www.oecd.org/en/data/datasets/income-and-wealth-distribution-database.html.

[125] OECD (n.d.), OECD Family Database, https://www.oecd.org/en/data/datasets/oecd-family-database.html (accessed on 11 September 2026).

[162] OECD (n.d.), OECD Health Statistics, https://data-explorer.oecd.org/s/5d4.

[163] OECD (n.d.), OECD Historial population data, https://data-explorer.oecd.org/s/5d5.

[160] OECD (n.d.), OECD National Accounts, https://data-explorer.oecd.org/s/5d3.

[24] OECD (n.d.), OECD National Accounts at a Glance (database), https://data-explorer.oecd.org/s/5c4 (accessed on 11 September 2026).

[159] OECD (n.d.), OECD Tax Benefit Model, https://data-explorer.oecd.org/s/5d2.

[123] OECD (n.d.), “PF1.3 Family cash benefits”, OECD Family Database, https://www.oecd.org/en/data/datasets/oecd-family-database.html (accessed on 14 September 2026).

[158] OECD (n.d.), Public and private social expenditure, https://data-explorer.oecd.org/s/5d1.

[16] Page, M. (2024), “New advances on an old question: Does money matter for children’s outcomes?”, Journal of Economic Literature, Vol. 62/3, pp. 891-947, https://doi.org/10.1257/jel.20231553.

[139] Pérez-Corral, A. and A. Moreno-Mínguez (2025), “Social spending, child deprivation and family structure: a multilevel study in 31 European countries”, International Review of Economics 2025 72:1, Vol. 72/1, pp. 8-, https://doi.org/10.1007/s12232-025-00483-0.

[17] Perry, B. (2021), Child Poverty in New Zealand: The demographics of child poverty, survey-based descriptions of life ‘below the line’ including the use of child-specific indicators, and international comparisons, Ministry of Social Develpment, Wellington.

[58] Petmesidou, M. et al. (2016), “Child poverty, youth (un)employment, and social inclusion”, p. 311, https://cup.columbia.edu/book/child-poverty-youth-unemployment-and-social-inclusion/9783838210025/ (accessed on 5 August 2025).

[110] Pilkauskas, N. et al. (2022), “The Effects of Income on the Economic Wellbeing of Families with Low Incomes: Evidence from the 2021 Expanded Child Tax Credit”, Working paper, No. 30533, N.B.E.R., http://www.nber.org/papers/w30533 (accessed on 15 September 2026).

[118] Price, D. and J. Song (2018), “The Long-Term Eﬀects of Cash Assistance”, IRS Working Papers, No. 621, Princeton University, Princeton, https://dataspace.princeton.edu/handle/88435/dsp01ng451m210 (accessed on 15 September 2026).

[29] Rapp, S. and O. Thévenon (2025), “Parenting on a budget: The cost of raising children in changing family configurations and its policy implications”, OECD Papers on Well-being and Inequalities, No. 36, OECD Publishing, Paris, https://doi.org/10.1787/bd0b594e-en.

[9] Schenck-Fontaine, A. and L. Panico (2019), “Many Kinds of Poverty: Three Dimensions of Economic Hardship, Their Combinations, and Children’s Behavior Problems”, Demography, https://doi.org/10.1007/s13524-019-00833-y.

[112] Schmidt, L. and S. Danziger (2012), “Filling holes in the safety net? Material hardship and subjective well-being among disability benefit applicants and recipients after the 1996 welfare reform”, Social Science Research, Vol. 41/6, pp. 1581-1597, https://doi.org/10.1016/J.SSRESEARCH.2012.06.001.

[60] Schulkind, L. and D. Sandler (2019), “The Timing of Teenage Births: Estimating the Effect on High School Graduation and Later-Life Outcomes”, Demography, Vol. 56/1, pp. 345-365, https://doi.org/10.1007/S13524-018-0748-6.

[113] Shaefer, H. and I. Gutierrez (2013), “The Supplemental Nutrition Assistance Program and material hardships among low-income households with children”, Social Service Review, Vol. 87/4, pp. 753-779, https://doi.org/10.1086/673999;REQUESTEDJOURNAL:JOURNAL:SSR;JOURNAL:JOURNAL:SSR;ISSUE:ISSUE:DOI.

[34] Thévenon, O. (2018), “Tackling child poverty in Canada”, OECD Social, Employment and Migration Working Papers, No. 220, OECD Publishing, Paris, https://doi.org/10.1787/dd4dcfa6-en.

[43] Thévenon, O. (2004), “Les enjeux pour l’emploi féminin de la Stratégie européenne pour l’emploi”, Revue de l’OFCE, Vol. 90/3, pp. 379-417, https://doi.org/10.3917/REOF.090.0379.

[80] Thévenon, O., C. Clarke and G. Simard-Duplain (2024), “The economic cost of childhood socio-economic disadvantage in Canada”, OECD Papers on Well-being and Inequalities, No. 25, OECD Publishing, Paris, https://doi.org/10.1787/7abf1837-en.

[69] Thévenon, O. et al. (2009), The costs of raising children and the effectiveness of supporting parenthood policies in European countries: A Literature Review.“ Report for the European Commission., INED, https://www.researchgate.net/publication/32221161_The_costs_of_raising_children_and_the_effectiveness_of_supporting_parenthood_policies_in_European_countries_A_Literature_Review (accessed on 7 September 2020).

[8] Thévenon, O. et al. (2018), “Child poverty in the OECD: Trends, determinants and policies to tackle it”, OECD Social, Employment and Migration Working Papers, No. 218, OECD Publishing, Paris, https://doi.org/10.1787/c69de229-en.

[22] UNICEF Innocenti (2023), Children and the Cost-of-living Crisis: How food and energy inflation has increased poverty in households with children in the European Union, UNICEF Innocenti Global Office of Research and Foresight, https://www.unicef.org/innocenti/reports/children-and-cost-living-crisis.

[165] van Buuren, S. (2018), Flexible Imputation of Missing Data, Second Edition, Chapman and Hall/CRC, https://stefvanbuuren.name/fimd/.

[147] Van Gestel, R. et al. (2023), “Improving Take-Up by Reaching Out to Potential Beneficiaries. Insights from a Large-Scale Field Experiment in Belgium”, Journal of Social Policy, Vol. 52/4, pp. 740-760, https://doi.org/10.1017/S004727942100088X.

[73] Van Lancker, W. (2023), “The Matthew Effect in Early Childhood Education and Care: How family policies may amplify inequalities”, The Oxford Handbook of Family Policy Over The Life Course, Oxford Academic, https://doi.org/10.1093/oxfordhb/9780197518151.013.36 (accessed on 26 November 2025).

[97] Van Lancker, W., J. Ghysels and B. Cantillon (2015), “The impact of child benefits on single mother poverty: Exploring the role of targeting in 15 European countries”, International Journal of Social Welfare, Vol. 24/3, pp. 210-222, https://doi.org/10.1111/IJSW.12140.

[136] Van Lancker, W. and N. Van Mechelen (2015), “Universalism under siege? Exploring the association between targeting, child benefits and child poverty across 26 countries”, Social Science Research, Vol. 50, pp. 60-75, https://doi.org/10.1016/J.SSRESEARCH.2014.11.012.

[40] Vandenbroucke, F., A. Hemerijck and B. Palier (2011), The EU Needs a Social Investment Pact, https://papers.ssrn.com/abstract=2991018 (accessed on 12 September 2025).

[68] Verbist, G. and W. Van Lancker (2016), “Horizontal and Vertical Equity Objectives of Child Benefit Systems: An Empirical Assessment for European Countries”, Social Indicators Research, Vol. 128/3, pp. 1299-1318, https://doi.org/10.1007/S11205-015-1080-9.

[94] Waldfogel, J. (2025), Child Benefits: A Smart Investment for America’s Future, Russell Sage Foundation, https://doi.org/10.7758/LIWL3023.

[132] Whiteford, P. (2010), “The Australian Tax-Transfer System: Architecture and Outcomes”, Economic Record, Vol. 86/275, pp. 528-544, https://doi.org/10.1111/j.1475-4932.2010.00634.x.

[65] Whiteford, P. and W. Adema (2007), “What Works Best in Reducing Child Poverty: A Benefit or Work Strategy?”, OECD Social, Employment and Migration Working Papers, No. 51, OECD Publishing, Paris, https://doi.org/10.1787/233310267230.

[72] Wood, J., K. Neels and J. Maes (2023), “A closer look at demand-side explanations for the Matthew effect in formal childcare uptake in Europe and Australia”, Journal of European Social Policy, Vol. 33/4, pp. 451-468, https://doi.org/10.1177/09589287231186068.

Annex 3.A. Child poverty rates by country

Copy link to Annex 3.A. Child poverty rates by country

Annex Figure 3.A.1. Children in single‑parent and large families are most at risk of income poverty
Copy link to Annex Figure 3.A.1. Children in single‑parent and large families are most at risk of income poverty

Child relative income poverty (50% of median earnings) by household composition, 2023 or latest available

Annex Figure 3.A.2. Children in single‑parent families are at higher risk of poverty
Copy link to Annex Figure 3.A.2. Children in single‑parent families are at higher risk of poverty

Annex 3.B. Sample description

Copy link to Annex 3.B. Sample description

The sample is based on available data for 32 OECD countries from 1997 to 2021, excluding Australia, Chile, Colombia, Costa Rica, Japan and Mexico due to extensive missing data across variables and over time. Detailed information on the indicator construction is provided in Annex Table 3.B.1 and summary statistics are presented in Annex Table 3.B.2. For variables that include imputed values, the statistics represent average values across the five datasets that result from the imputation process. For further information on the imputations, see Annex 3.C. The regression analyses are restricted to the period 2002-2021. Data for the earlier years (1997 to 2001) serve to incorporate time lags.

Indicator	Definition	Data source
Outcome variables
Child relative poverty rate	The share of children living in households with equivalised disposable incomes that are below 50% of the national median equivalised disposable income, denoted in nominal terms (current prices)	OECD Income Distribution Database (OECD, n.d.[19]),
Child anchored child poverty rate	The share of children living in households with equivalised disposable incomes that are below 50% of the national median equivalised disposable income in 2005, adjusted for inflation, i.e. denoted in real terms (constant prices)	OECD Income Distribution Database (OECD, n.d.[19]),
Severe material deprivation rate, ages 0‑17	The share of children living in households that experience severe material deprivation. As defined by Eurostat (2026[156]), material deprivation refers to a state of economic strain and durables. The severe material deprivation rate is defined as the enforced inability (rather than the choice not to do so) to afford at least four of the following items considered by most people to be desirable or even necessary to lead an adequate life: 1) to pay their rent, mortgage or utility bills; 2) to keep their home adequately warm; 3) to face unexpected expenses; 4) to eat meat or proteins regularly; 5) to go on holiday; 6) a television set; 7) a washing machine; 8) a car; 9) a telephone	Eurostat Population and social conditions (Eurostat, n.d.[157])
Policy variables
Spending on ALMP	The variable refers to the share of public and mandatory private and voluntary private expenditures on active labour market programmes in the total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The variable is denoted in per capita USD PPP and constant 2015 prices.	OECD Social Expenditure Database (OECD, n.d.[158])
ECEC enrolment 0‑2	The per cent of children aged 0 to 2 enrolled in formal care institutions. Some values have been imputed.	OECD Family Database – Indicator PF3.2 (OECD, n.d.[125])
ECEC enrolment 3‑5	The per cent of children aged 3 to 5 enrolled in early childhood education and care services. Some values have been imputed.	OECD Family Database – Indicator PF3.2 (OECD, n.d.[125])
Spending on ECEC	The variable refers to the share of public and mandatory private and voluntary private expenditures on ECEC services in the total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The variable is denoted in per capita USD PPP and constant 2015 prices.	OECD Social Expenditure Database (OECD, n.d.[158])
Spending on health services	The variable refers to the share of public and mandatory private and voluntary private expenditures on health in the total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The variable is denoted in per capita USD PPP and constant 2015 prices	OECD Social Expenditure Database (OECD, n.d.[158])
Other social spending in kind	The indicator comprises the public and mandatory private and voluntary private expenditures paid in kind for incapacity-related, housing and other social policy areas as a share of total social expenditure (cash and in-kind in the areas of families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The variable is denoted in per capita USD PPP and constant 2015 prices. The indicator excludes from the numerator spending on active labour market programmes, families and health (they enter as separate policy variables), unemployment (countries do not report in-kind benefits in this area), survivors (they mainly cover in-kind benefits related to the funeral and are thus less relevant to child outcomes) and old-age benefits (outside the scope).	OECD Social Expenditure Database (OECD, n.d.[158])
Effective tax rates on taking up work – single‑parent family	This indicator measures the financial disincentives to participate in the labour market. It calculates the proportion of earnings in the new job that are lost to either higher taxes and social contributions, or lower benefit entitlements when a jobless parent with two children aged 4 and 6 takes up full-time employment at 67% of the average national wage after five years of claiming unemployment benefits. Some values have been imputed.	OECD Tax-Benefit Model (OECD, n.d.[159])
Effective tax rates on increasing working hours – single‑parent family	This indicator measures the percentage of additional earnings that is lost to either higher taxes and social contributions, or lower benefits when a single parent with two children aged 4 and 6 who earns 67% of the average national wage increases their working hours from 50% to 100%. Some values have been imputed.	OECD Tax-Benefit Model (OECD, n.d.[159])
Effective tax rates on taking up work – two‑parent family	This indicator measures the financial disincentives to participate in the labour market. It calculates the proportion of earnings in the new job that are lost to either higher taxes or lower benefit entitlements when the second earner in a couple family with two children aged 4 and 6 takes up full-time employment at 50% of the average national wage after five years of claiming unemployment benefits. The other parent is always assumed to work full-time at 50% of the average national wage. Some values have been imputed.	OECD Tax-Benefit Model (OECD, n.d.[159])
Effective tax rates on increasing working hours – two‑parent family	This indicator measures the percentage of additional earnings that is lost to either higher taxes or lower benefits when the second earner in a couple family with two children aged 4 and 6 who earns 50% of the average national wage increases their working hours from 50% to 100%. The other parent is always assumed to work full-time at 50% of the average national wage. Some values have been imputed.	OECD Tax-Benefit Model (OECD, n.d.[159])
Volume of redistribution	This indicator is calculated as the average amount per equivalent household member of public transfers received (public social security, excluding employment-related social insurance transfers and private transfers received) minus taxes on income and wealth paid and of contributions paid to public social security schemes by working-age individuals, expressed as a percentage of GDP per capita. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]) and OECD National Accounts (OECD, n.d.[160])
Average size of taxes	The average taxes per equivalent household member paid by working-age individuals, as a share of average disposable income per equivalent household member of working-age individuals. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Average size of benefits	The average public transfers (public social security) per equivalent household member received by working-age individuals, as a share of average disposable income per equivalent household member of working-age individuals. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Share of taxes paid by the bottom 30%	The taxes per equivalent household member paid by working-age individuals in the bottom 30% of post-transfer, equivalised household income, as a share of taxes per equivalent household member paid by all working-age individuals. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Share of benefits received by the bottom 30%	The public transfers (public social security) per equivalent household member received by working-age individuals in the bottom 30% of pre‑transfer income, as a share of public transfers per equivalent household member received by all working-age individuals. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Cash transfers to families	The variable refers to the share of public and mandatory private and voluntary private expenditures allocated to families in cash in the total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The variable is denoted in per capita USD PPP and constant 2015 prices	OECD Social Expenditure Database (OECD, n.d.[158])
Cash transfers to families during early childhood (ages 0‑5)	This indicator approximates the public and mandatory private and voluntary private expenditures that is allocated in cash to families with children aged 0‑5 as a share of total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The indicator is denoted in per capita USD PPP and constant 2015 prices. It is constructed by multiplying the indicator “Cash transfers to families” described above with age‑spending profiles. The latter capture the average family cash benefits for a child at ages 0‑5 as a share of the average cash benefits for children at ages 0‑17 and contain imputed values for some years.	OECD Social Expenditure Database (OECD, n.d.[158]) and OECD Family Database – Indicator PF1.6 (OECD, n.d.[125])
Cash transfers to families during middle childhood (ages 6‑11)	This indicator approximates the public and mandatory private and voluntary private expenditures that is allocated in cash to families with children aged 6‑11 as a share of total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The indicator is denoted in per capita USD PPP and constant 2015 prices. It is constructed by multiplying the indicator “Cash transfers to families” described above with age‑spending profiles. The latter capture the average family cash benefits for a child at ages 6‑11 as a share of the average cash benefits for children at ages 0‑17 and contain imputed values for some years.	OECD Social Expenditure Database (OECD, n.d.[158]) and OECD Family Database – Indicator PF1.6 (OECD, n.d.[125])
Cash transfers to families during adolescence (ages 12‑17)	This indicator approximates the public and mandatory private and voluntary private expenditures that is allocated in cash to families with children aged 12‑17 as a share of total social expenditure (including spending on families, health, survivor and incapacity-related benefits, housing, active labour market programmes, unemployment, and other social policy areas, but excluding spending on old age and pensions). The indicator is denoted in per capita USD PPP and constant 2015 prices. It is constructed by multiplying the indicator “Cash transfers to families” described above with age‑spending profiles. The latter capture the average family cash benefits for a child at ages 12‑17 as a share of the average cash benefits for children at ages 0‑17 and contain imputed values for some years.	OECD Social Expenditure Database (OECD, n.d.[158]) and OECD Family Database – Indicator PF1.6 (OECD, n.d.[125])
RED index	The RED index measures the progressivity of taxes and benefits weighted by the volume of redistribution for a hypothetical set of households, following the methodology outlined in the OECD tax-benefit simulation model (2024[161]). For every pair of country $i$ and time $t$ , the formula is ${R E D i n d e x}_{i t} = I_{i t} * K_{i t}$ where $I_{t} = t_{i t} / (1 - t_{i t})$ measures the size of redistribution, or the incidence, and $t_{i t} = \sum_{j} ({t a x e s}_{i t j} - {b e n e f i t s}_{i t j}) / \sum_{j} {g r o s s i n c o m e}_{i t j}$ is the total of net taxes paid as a share of total gross income of across households $J$ . $K_{i t} = {C o n c e n t r a t i o n i n d e x}_{i t} ({t a x e s}_{i t j} - {b e n e f i t s}_{i t j}) - {G i n i i n d e x}_{i t} ({g r o s s i n c o m e}_{i t j})$ is the Kakwani index and the formula for the concentration index takes the form ${C o n c e n t r a t i o n i n d e x}_{i t} = 2 * \sum_{j} (({t a x e s}_{i t j} - {b e n e f i t s}_{i t j}) * R_{j}) / \sum_{j} ({t a x e s}_{i t j} - {b e n e f i t s}_{i t j}) - 1$ with the rank of household $j$ being calculated as $R_{i t j} = (j * 0.5) / J$ where $J$ represents the total number of households. The RED index yields values between -1 and 1, with higher values indicating a more progressive tax schedule. A tax schedule that is exactly proportional to gross income has a RED index of 0 as the concentration of net tax liabilities would be exactly the same as the distribution of gross income, which results in a Kakwani index of 0. A tax schedule that generates larger tax liabilities as a percentage of gross income for higher-income groups will result in a positive RED index. The incidence – or size – weighs the redistributive effect by the volume of redistribution; as a consequence, highly progressive benefits will not change the RED index by a lot if the benefit size is small. The notation of taxes is simplified and includes income reductions due to both taxes and social contributions. The set of hypothetical households is comprised of observations across 7 types of families, each with 0‑3 children (28 household types in total). The seven family types are: 1) a single earner working full-time, 2) a single earner working part time (50%), 3) a single earner who is unemployed, 4) a couple where the first earner works full-time and the second earner is employed full-time at 67% of the average wage, 5) a couple where the first earner works full-time and the second earner is employed part-time (50%) at 67% of the average wage, 6) a couple where the first earner works full-time and the second earner does not work, and 7) a couple where both are unemployed The earnings of the first earner, where they are employed, range from the national minimum wage (or the equivalent of the OECD average where there is none) to 1.5 times the national median equivalent disposable income. Every percentile of the national average wage within that range represents one observation of a hypothetical household. In addition, observations for the two household types where all adults are unemployed are replicated to represent 10% of the other five family types.	OECD Tax-Benefit Model (OECD, n.d.[159])
Income transfers for 1‑3 children	This set of three indicators measures the additional benefits (net of taxes and social contributions paid) that a hypothetical set of low- to middle‑income families with 1 child (or 2 or 3 children, respectively) receive compared to the same set of families without any children, as a share of their gross income. Mathematically, for every country $i$ and time $t$ , the indicator for families with the number of children $a \in {1, 2, 3}$ is $\sum_{j, c = a} ({b e n e f i t s}_{i t j c} - {t a x e s}_{i t j c}) - \sum_{j, c = 0} ({b e n e f i t s}_{i t j c} - {t a x e s}_{i t j c})) / \sum_{j} {g r o s s i n c o m e}_{i t j}$ where $c$ is the number of children of household $j$ . The sample of hypothetical households contains 7 types of families, each with 0‑3 children (28 household types in total). The seven household types are: a single earner working full-time, a single earner working part time (50%), a single earner who is unemployed, a couple where the first earner works full-time and the second earner is employed full-time at 67% of the average wage, a couple where the first earner works full-time and the second earner is employed part-time (50%) at 67% of the average wage, a couple where the first earner works full-time and the second earner does not work, and a couple where both are unemployed The earnings of the first earner, where they are employed, range from the national minimum wage (or the equivalent of the OECD average where there is none) to 1.5 times the national median equivalent disposable income. Every percentile of the national average wage within that range represents one observation of a hypothetical household. Households where all parents are unemployed are represented with one observation.	OECD Tax-Benefit Model (OECD, n.d.[159])
Tax breaks for families	The indicator captures the volume of tax breaks for families in per cent of GDP, divided by the share of the population under age 20. Tax exemptions for families include income from child benefits that is not included in the tax base, child tax allowances (amounts for children that are deducted from gross income and are not included in taxable income), and child tax credits (amounts that are deducted from the tax liability).	OECD Family Database – Indicator PF1.1 (OECD, n.d.[125])
Child poverty drivers
Total social expenditure (TS)	Total social spending refers to countries’ public, private mandatory and private voluntary expenditures in-cash and in-kind in eight different policy areas: health, incapacity-related benefits, unemployment, active labour market programmes, family, housing, survivors, and other social policy areas (e.g. social assistance). It excludes spending on old age. It is denoted in constant USD PPP (base 2015 = 100) and in logarithmic form.	OECD Social Expenditure Database, (OECD, n.d.[158])
Adolescent fertility rate	The number of births per 1 000 women aged 15‑19 at childbirth	OECD Family Database – Indicator SF2.3 (OECD, n.d.[125])
Share of single‑parent families	The percentage share of households with children that are headed by a single parent among all households with children. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Average number of children per household	The number of children divided by the number of households in a country. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Mean poverty gap before taxes and transfers	The mean poverty gap before taxes and transfers in per cent, i.e. the difference between the average equivalised pre‑tax income of individuals living in a working-age household below the relative poverty line and the relative poverty line as a share of the relative poverty line. Some values have been imputed.	OECD Income Distribution Database (OECD, n.d.[19]),
Total social spending drivers
GDP per capita	GDP per capita net of total social spending per capita (as defined in “total social expenditure” above), denoted in USD PPP in constant prices and in logarithmic form.	OECD National Accounts (OECD, n.d.[160])
Life expectancy	Years of life expectancy at birth, denoted in logarithmic form.	OECD Health Statistics (OECD, n.d.[162])
Dependency ratio	The population aged less than 20 years or 65 years or over as a ratio of the population aged 20‑64.	OECD Historical population data (OECD, n.d.[163])

Variable	Unit	Standard deviation	Mean	Minimum	Maximum
Child relative poverty rate	Percentage of children	5.16	12.18	2.10	27.46
Child anchored poverty rate	Percentage of children	6.77	9.78	0.42	41.23
Severe material deprivation rate among children	Percentage of children	10.48	9.26	0.20	66.70
ALMP	Percentage of social expenditure (excl. old age spending)	2.33	3.33	0.00	19.52
ECEC enrolment, ages 0‑2	Percentage of children	18.12	30.35	0.19	76.20
ECEC enrolment, ages 3‑5	Percentage of children	17.74	80.69	31.71	100.10
Spending on ECEC	Percentage of social expenditure (excl. old age spending)	2.26	3.96	0.00	10.66
Effective tax rates on increasing working hours – two‑parent family	Percentage of additional earnings	20.13	39.14	6.32	119.93
Effective tax rates on increasing working hours – single‑parent family	Percentage of additional earnings	21.23	51.88	8.00	100.00
Effective tax rates on taking up work – two‑parent family	Percentage of additional earnings	20.90	44.87	6.30	105.80
Effective tax rates on taking up work – single‑parent family	Percentage of additional earnings	20.03	57.12	‑0.20	100.00
Health services	Percentage of social expenditure (excl. old age spending)	9.37	44.67	23.33	81.03
Other in-kind spending	Percentage of social expenditure (excl. old age spending)	3.49	5.44	0.19	17.97
Tax breaks for families	Percentage of GDP per population share under age 20	1.16	0.92	0.00	5.02
Volume of redistribution	Percentage of GDP	0.05	‑0.05	‑0.18	0.08
Average size of taxes	Percentage of average disposable income	10.80	28.42	0.02	56.55
Average size of benefits	Percentage of average disposable income	6.18	16.47	3.18	34.82
Taxes paid by bottom 30%	Percentage of total taxes paid	3.52	9.34	1.33	26.77
Benefits received by bottom 30%	Percentage of total benefits received	10.49	33.70	12.00	58.00
Income transfers for 1 child	Index	4.13	8.95	‑0.54	20.61
Income transfers for 2 children	Index	6.56	16.30	0.04	40.67
Income transfers for 3 children	Index	11.84	26.22	0.06	73.48
RED index	Index	0.04	0.12	0.01	0.21
Cash transfers to families	Percentage of social expenditure (excl. old age spending)	4.46	8.72	0.25	25.01
Cash transfers to families during early childhood (ages 0‑5)	Percentage of social expenditure (excl. old age spending)	2.66	4.21	0.11	18.78
Cash transfers to families during middle childhood (ages 6‑11)	Percentage of social expenditure (excl. old age spending)	1.23	2.19	0.11	6.19
Cash transfers to families during adolescence (ages 12‑17)	Percentage of social expenditure (excl. old age spending)	1.26	2.12	0.10	6.19
Social expenditure	USD PPP (base year = 2015)	3 448.01	6 293.26	626.84	18 794.87
Social expenditure, logged	Log of USD PPP (base year = 2015)	0.65	8.57	6.44	9.84
Adolescent fertility rate, per 1 000	Ratio	8.71	12.85	0.40	51.30
Share of single‑parent families among all families	Percentage of households	3.88	7.58	1.31	18.91
Average number of children per household	Ratio	0.19	0.54	0.30	1.67
Mean poverty gap before taxes and transfers	Percentage of the relative poverty line (50% of equivalised median disposable income)	8.39	61.11	35.57	80.62
GDP per capita	USD PPP (base year = 2020)	15 585.90	39 239.12	10 923.12	11 0 467.82
GDP per capita, logged	Log of USD PPP (base year = 2020)	0.38	10.51	9.30	11.61
Life expectancy at birth	Years	3.13	79.07	69.70	84.20
Life expectancy at birth, logged	Log of years	0.04	4.37	4.24	4.43
Dependency ratio	Ratio	7.13	66.03	48.15	92.67

Note: These summary statistics are based on available data for 32 OECD countries excluding Australia, Chile, Columbia, Costa Rica, Japan and Mexico from 1997 to 2021. For variables that include imputed values, the statistics represent average values across the five datasets that result from the imputation process. For further information on the imputations, see Annex 3.C.

Annex 3.C. Imputation of missing observations

Copy link to Annex 3.C. Imputation of missing observations

Many indicators don’t show a complete set of values and missing values are not at random. To address these issues, missing data are imputed with two methods. First, Stineman interpolation is used between observed values for indicators that are unlikely to change substantially over time and where only few observations are missing. These are: tax breaks for families, cash transfers to families at different ages (ages 0-5, 6-11 and 12-17), and the average number of children per household. Imputations were implemented via the R package imputeTS (Moritz and Bartz-Beielstein, 2017[164]).

In addition, a multiple imputation approach (e.g. van Buuren (2018[165])) is used for the following variables: ECEC enrolment at ages 0-2 and 3-5, effective tax rates on increasing working hours and on taking up work, the volume of redistribution, taxes paid by the bottom 30%, average size of taxes, benefits received by the bottom 30%, average size of benefits, the mean poverty gap before taxes and transfers, the share of single parent families, tax breaks for families and the average number of children per household. This non-parametric approach uses the distribution of the observed variables to estimate multiple possible values for the missing observation, therefore allowing to account for the uncertainty around the true value for which data is missing. With this approach, the available statistical information and the number of observations that enter the analysis are maximised, which is especially important in this small-sample cross-country setting. The resulting estimates are consistent, asymptotically normal and close to being asymptotically efficient (Kropko et al., 2014[166]).

In practice, the multiple imputation model replaces the missing observations using predictive mean matching. As a first step, this process estimates a predicted value for the missing observation using the information in the remaining dataset and selects a set of 5 complete observations with predicted values that are closest in absolute distance to the predicted value for the missing observation. Assuming that the missing observation is based on the same distribution as these 5 observations, one of these observations is drawn at random to replace the missing observation in the dataset. This process is repeated 5 times, forming 5 separate datasets with imputed data for the missing observation on which the statistical analyses are performed. The resulting coefficients are then pooled, while adjusting the standard errors for the uncertainty that is produced by the imputation method (van Buuren, 2018[165]).

Annex 3.D. Regression results

Copy link to Annex 3.D. Regression results

In the regression estimations presented below, policy variables and the two main drivers – the log of total social expenditure per capita and the log of GDP per capita – are standardised using all available observations, including imputed values for variables where applicable, for the 32‑country sample over the period 1997-2021. First differenced and lagged variables are constructed after standardisation, and the regression analysis is restricted to the period 2002-2021. For specifications in which the dependent variable is available for a shorter period, the estimation sample is reduced correspondingly.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑0.806**	‑0.886*	‑0.841**	‑0.703*	‑0.978***	‑0.678**	‑0.588*
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑0.308	‑0.309	‑0.317	‑0.311	‑0.181	‑0.153	‑0.431
ECEC enrolment 0‑2 x TS	‑0.897		‑0.994	‑0.784	‑0.85	‑0.705	‑0.912
ECEC enrolment 0‑2 x TS	‑0.504		‑0.74	‑0.736*	‑0.355	‑0.594	‑0.409
ECEC enrolment 3‑5 x TS	‑1.52***		‑1.865***	‑1.453***	‑1.226***	‑1.448***	‑1.317***
ECEC enrolment 3‑5 x TS	0.004		0.031	0.163	0.297	‑0.205	‑0.143
Spending on ECEC x TS		0.233	0.931**
Spending on ECEC x TS		0.046	0.137
ECEC enrolment 0‑2 x spending on ECEC			0.191
ECEC enrolment 3‑5 x spending on ECEC			‑0.407*
Lower ETR single: new job x TS				0.83***
Lower ETR single: new job x TS				0.017
Lower ETR single: additional hours x TS				‑0.238
Lower ETR single: additional hours x TS				0.229
Lower ETR couple: new job x TS					0.706**
Lower ETR couple: new job x TS					‑0.281
Lower ETR couple: additional hours x TS					‑0.035
Lower ETR couple: additional hours x TS					0.836***
Spending on health services x TS						0.166
Spending on health services x TS						1.148***
Other social spending in kind x TS	‑0.238	‑0.188	‑0.181	‑0.086	0.218	‑0.531	‑0.426
Other social spending in kind x TS	‑0.62*	‑1.038***	‑0.5	‑0.668*	‑0.604	0.04	‑0.352
Tax breaks for families x TS							‑0.274
Tax breaks for families x TS							0.429*
Volume of redistribution x TS	‑0.112	0.11	0.007	‑0.049	‑0.432	‑0.073	‑0.22
Volume of redistribution x TS	‑0.419	‑0.388	‑0.688**	‑0.494	‑0.376	‑0.471	‑0.414
Share of taxes paid by the bottom 30% x TS	0.192	‑0.238	0.024	0.266	0.581	0.285	0.05
Share of taxes paid by the bottom 30% x TS	‑0.11	‑0.356	‑0.265	‑0.047	0.515	‑0.171	‑0.174
Share of benefits received by the bottom 30% x TS	‑0.685	‑1.348	‑0.869	‑0.708	‑0.885	‑0.475	‑0.818
Share of benefits received by the bottom 30% x TS	0.48	0.985**	0.295	0.497	0.705	0.048	0.368
In-cash spending on families x TS	‑1.461***	‑1.187**	‑1.394***	‑1.371***	‑1.4***	‑1.597***	‑1.453***
In-cash spending on families x TS	‑0.741*	‑0.398	‑0.726	‑1.028**	‑1.121**	‑0.676*	‑0.724
Total social expenditure per capita (logged)	‑2.94***	‑2.656**	‑2.589***	‑3.452***	‑3.585***	‑3.097***	‑3.216***
Adolescent fertility rate (per 1 000)	0.113***	0.152***	0.09**	0.091**	0.097**	0.133***	0.124***
Share of single‑parent families	0.281	0.318	0.269	0.232	0.285*	0.402*	0.264
Average number of children per household	1.099	‑1.876	‑0.363	5.443	3.965	3.891	‑2.566
Mean poverty gap	0.018	‑0.014	‑0.015	0.069	0.024	0.02	0.011
Adjusted R²	0.409	0.358	0.441	0.443	0.469	0.433	0.415
Number of observations	478	478	478	478	461	478	478
Number of countries	32	32	32	32	32	32	32
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Note: Coefficients in bold are the ones used for the value for money calculations. “x” signifies an interaction between two variables. “x TS” means the interaction between the policy variable in the line above and the log of per capita total social spending. The log of total social spending per capita is standardised along with the policy variables to facilitate the interpretation of results.

	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑0.908**	‑0.968**	‑1.066***	‑1.379***	‑1.304***	‑1.196***	‑0.789*	‑1.051**	‑0.967**
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑0.205	‑0.228	‑0.111	‑0.128	‑0.153	‑0.004	0.297	0.324	0.367
ECEC enrolment 0‑2 x TS	‑0.708	‑0.879	‑0.993	‑0.783	‑0.736	‑0.489	‑0.914	‑0.987	‑1.042*
ECEC enrolment 0‑2 x TS	‑0.293	‑0.538	‑0.013	‑0.067	‑0.131	0.107	‑0.326	‑0.476	‑0.343
ECEC enrolment 3‑5 x TS	‑1.342***	‑1.515***	‑1.544***	‑1.15***	‑1.033**	‑1.05***	‑1.618***	‑1.519***	‑1.346***
ECEC enrolment 3‑5 x TS	0.002	0.085	0.002	‑0.104	‑0.126	0.158	0.025	‑0.28	‑0.192
Other social spending in kind x TS	‑0.378	‑0.626	0.264	0.376	0.269	‑0.182	‑0.809	‑1.158*	‑1.061
Other social spending in kind x TS	‑0.637*	‑0.278	‑0.611*	‑0.688*	‑0.749**	‑0.42	‑0.27	0.247	0.092
Volume of redistribution x TS			‑0.228	‑0.302	‑0.17		0.404	0.498	0.507
Volume of redistribution x TS			‑0.322	‑0.373	‑0.392		‑0.532	‑0.69	‑0.781
Size of benefits	‑1.248**	‑0.136
Size of taxes	‑0.334	‑0.241
Share of taxes paid by the bottom 30% x TS	‑0.134		0.301	0.335	0.273		‑0.071	0.46	0.396
Share of taxes paid by the bottom 30% x TS			‑0.042	0.05	0.071		‑0.38	‑0.412	‑0.354
Share of benefits received by the bottom 30% x TS	‑1.36		‑1.062	‑0.938	‑0.881		‑0.29	0.318	0.179
Share of benefits received by the bottom 30% x TS			0.737	0.489	0.548		0.617	0.053	0.112
Size of benefits x share of benefits received by the bottom 30%	‑1.12***
Size of taxes x share of taxes received by the bottom 30%	‑0.233*
Income transfers for 1 child x TS			‑0.733**
Income transfers for 1 child x TS			‑1.258***
Income transfers for 2 children x TS				‑1.142***
Income transfers for 2 children x TS				‑0.803***
Income transfer for 3 children x TS					‑1.292***
Income transfer for 3 children x TS					‑0.995***
RED index x TS						‑1.37***
RED index x TS						‑1.128***
In-cash spending on families x TS	‑1.384***	‑1.634***				‑1.212***
In-cash spending on families x TS	‑0.754*	‑0.989**				‑0.57
In-cash spending on families, ages 0‑5 x TS							‑1.127
In-cash spending on families, ages 0‑5 x TS							‑1.059
In-cash spending on families, ages 6‑11 x TS								‑1.582***
In-cash spending on families, ages 6‑11 x TS								0.113
In-cash spending on families, ages 12‑17 x TS									‑1.304***
In-cash spending on families, ages 12‑17 x TS									0.207
Total social expenditure per capita (logged)	‑2.926***	‑3.719***	‑3.032***	‑2.843***	‑2.757***	‑2.982***	‑3.068*	‑3.028*	‑2.73*
Adolescent fertility rate (per 1 000)	0.104***	0.124***	0.162***	0.145***	0.135***	0.147***	0.132***	0.12***	0.14***
Share of single‑parent families	0.276	0.248	0.341*	0.388*	0.367*	0.31*	‑0.114	0.062	0.063
Average number of children per household	1.723	0.054	‑0.95	‑1.475	0.755	5.002	5.58	6.021	5.246
Mean poverty gap	0.094	0.049	0.034	0.049	0.053	0.027	0.059	0.09	0.1*
Adjusted R²	0.469	0.391	0.456	0.463	0.461	0.464	0.42	0.445	0.436
Number of observations	478	478	465	465	465	465	333	333	333
Number of countries	32	32	32	32	32	32	28	28	28
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑2.021***	‑2.109**	‑1.665**	‑2.042***	‑2.567***	‑2.05***	‑1.738**
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	1.115**	1.377***	0.765	1.05**	0.977	1.024**	0.835
ECEC enrolment 0‑2 x TS	‑1.786*		‑1.813	‑1.703	‑1.819*	‑2.047*	‑1.718*
ECEC enrolment 0‑2 x TS	‑0.892		‑1.349***	‑1.186**	‑0.947	‑1.056**	‑0.966*
ECEC enrolment 3‑5 x TS	0.744		‑0.858	0.709	1.46**	1.126**	0.974*
ECEC enrolment 3‑5 x TS	0.401		0.524	‑0.004	0.548	0.25	0.156
Spending on ECEC x TS		0.798	2.019*
Spending on ECEC x TS		‑0.257	0.44
ECEC enrolment 0‑2 x spending on ECEC			0.28
ECEC enrolment 3‑5 x spending on ECEC			‑1.717**
Lower ETR single: new job x TS				‑1.126***
Lower ETR single: new job x TS				‑0.143
Lower ETR single: additional hours x TS				0.131
Lower ETR single: additional hours x TS				‑0.42
Lower ETR couple: new job x TS					‑0.385
Lower ETR couple: new job x TS					‑0.467
Lower ETR couple: additional hours x TS					1.293**
Lower ETR couple: additional hours x TS					0.709*
Spending on health services x TS						‑4.146***
Spending on health services x TS						0.267
Other social spending in kind x TS	‑2.922**	‑3.463***	‑2.629**	‑2.635**	‑2.666**	‑4.588***	‑3.485**
Other social spending in kind x TS	2.679**	2.845**	2.468**	2.408**	3.152***	3.256***	3.429**
Tax breaks for families x TS							‑0.229
Tax breaks for families x TS							0.81*
Volume of redistribution x TS	‑0.125	‑0.457	0.085	‑0.513	‑0.521	0.307	‑0.267
Volume of redistribution x TS	0.103	0.094	‑0.593	‑0.284	0.097	‑0.726	0.188
Share of taxes paid by the bottom 30% x TS	1.692*	1.941**	1.106	1.923**	2.628***	1.763**	1.416
Share of taxes paid by the bottom 30% x TS	‑0.066	‑0.44	‑0.426	‑0.284	0.603	‑0.089	‑0.226
Share of benefits received by the bottom 30% x TS	‑3.048**	‑3.328**	‑2.862**	‑2.89***	‑3.737**	‑1.117	‑3.022**
Share of benefits received by the bottom 30% x TS	1.559	1.706	1.018	1.71*	1.734	0.809	1.392
In-cash spending on families x TS	‑0.152	‑0.143	0.008	‑0.06	‑0.342	‑1.113	‑0.246
In-cash spending on families x TS	‑0.113	‑0.266	‑0.04	‑0.002	‑0.23	0.687	‑0.035
Total social expenditure per capita (logged)	‑7.304***	‑6.042***	‑5.778***	‑7.297***	‑6.745***	‑8.825***	‑7.584***
Adolescent fertility rate (per 1 000)	0.12	0.058	0.019	0.093	0.092	0.134	0.123
Share of single‑parent families	0.816**	0.924**	0.865**	0.778**	0.965***	0.708**	0.802**
Average number of children per household	15.809	14.148	16.222	15.189	6.941	23.727**	12.457
Mean poverty gap	0.727***	0.655***	0.588***	0.694***	0.634***	0.602***	0.702***
Adjusted R²	0.675	0.667	0.709	0.689	0.688	0.71	0.678
Number of observations	411	411	411	411	399	411	411
Number of countries	30	30	30	30	30	30	30
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes

	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑1.379**	‑1.071**	‑2.501***	‑2.569***	‑2.344***	‑1.788**	‑2.93***	‑2.184**	‑2.212**
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	0.854	0.68	1.349**	1.199**	1.039*	0.565	2.281***	1.55**	1.508**
ECEC enrolment 0‑2 x TS	‑0.786	‑1.712	‑2.017*	‑2.177**	‑2.431*	‑2.633*	‑2.109*	‑2.473**	‑2.555**
ECEC enrolment 0‑2 x TS	‑0.872**	‑1.395***	‑0.637	‑0.761	‑0.677	‑0.926**	‑0.773	‑0.175	‑0.227
ECEC enrolment 3‑5 x TS	0.504	0.526	0.771	0.963*	0.84	0.479	0.98	0.689	0.523
ECEC enrolment 3‑5 x TS	0.594	0.744	‑0.135	0.006	0.187	0.784	‑1.165*	‑0.705	‑0.756
Other social spending in kind x TS	‑3.169***	‑4.2***	‑2.69***	‑2.57**	‑2.345**	‑3.506***	‑3.461**	‑1.725	‑1.947
Other social spending in kind x TS	2.333**	3.394***	2.953***	2.79***	2.636**	2.763***	2.581*	1.219	1.423
Volume of redistribution x TS			‑0.561	‑0.326	‑0.353		1.47	1.034	0.947
Volume of redistribution x TS			0.368	0.262	0.296		‑0.108	‑0.393	‑0.315
Size of benefits	1.524**	3.528***
Size of taxes	‑0.409	0.512
Share of taxes paid by the bottom 30% x TS	0.883		2.221**	2.538***	2.69***		2.422***	2.386***	2.447***
Share of taxes paid by the bottom 30% x TS			‑0.255	0.101	0.261		‑0.617	‑0.768	‑0.803
Share of benefits received by the bottom 30% x TS	‑1.375		‑4.036***	‑3.278**	‑3.271**		‑4.265**	‑4.454**	‑4.232*
Share of benefits received by the bottom 30% x TS			1.795	1.578	1.652		2.465**	2.212**	2.072*
Size of benefits x share of benefits received by the bottom 30%	‑2.526***
Size of taxes x share of taxes received by the bottom 30%	‑0.185
Income transfers for 1 child x TS			‑0.815
Income transfers for 1 child x TS			‑0.171
Income transfers for 2 children x TS				‑0.42
Income transfers for 2 children x TS				0.144
Income transfer for 3 children x TS					0.141
Income transfer for 3 children x TS					‑0.006
RED index x TS						‑0.12
RED index x TS						0.189
In-cash spending on families x TS	‑0.491	‑0.988				0.127
In-cash spending on families x TS	‑0.196	‑0.333				0.152
In-cash spending on families, ages 0‑5 x TS							‑1.669
In-cash spending on families, ages 0‑5 x TS							‑0.507
In-cash spending on families, ages 6‑11 x TS								0.867
In-cash spending on families, ages 6‑11 x TS								‑0.34
In-cash spending on families, ages 12‑17 x TS									0.526
In-cash spending on families, ages 12‑17 x TS									‑0.359
Total social expenditure per capita (logged)	‑7.007***	‑9.146***	‑5.778***	‑6.339***	‑6.717***	‑7.489***	‑12.238***	‑12.051***	‑12.34***
Adolescent fertility rate (per 1 000)	0.268***	0.283***	0.144	0.106	0.083	0.1	0.314**	0.237*	0.218*
Share of single‑parent families	0.598*	0.514	0.922***	0.942***	0.934***	0.521	0.056	0.165	0.19
Average number of children per household	5.912	8.029	2.94	8.3	11.287	22.669*	13.061	7.04	8.752
Mean poverty gap	0.417***	0.38***	0.766***	0.734***	0.713***	0.752***	0.83***	0.821***	0.816***
Adjusted R²	0.771	0.696	0.681	0.68	0.679	0.642	0.727	0.727	0.727
Number of observations	411	411	402	402	402	402	292	292	292
Number of countries	30	30	30	30	30	30	26	26	26
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑1.628*	‑2.399**	‑1.829*	‑1.966**	‑2.05*	‑1.846*	‑1.522
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	1.451*	1.733**	1.382	1.491**	2.077**	1.051	1.242
ECEC enrolment 0‑2 x TS	‑0.342		‑0.928	‑0.176	‑0.653	‑0.351	‑0.48
ECEC enrolment 0‑2 x TS	0.115		‑0.191	0.101	0.143	‑0.265	0.158
ECEC enrolment 3‑5 x TS	‑0.256		‑0.613	‑0.841	0.029	0.326	0.002
ECEC enrolment 3‑5 x TS	2.036***		2.16***	2.356***	1.97**	2.289***	1.802**
Spending on ECEC x TS		0.623	1.062
Spending on ECEC x TS		0.363	0.25
ECEC enrolment 0‑2 x spending on ECEC			0.136
ECEC enrolment 3‑5 x spending on ECEC			‑0.36
Lower ETR single: new job x TS				‑0.505
Lower ETR single: new job x TS				‑0.275
Lower ETR single: additional hours x TS				‑1.342**
Lower ETR single: additional hours x TS				0.577
Lower ETR couple: new job x TS					‑1.552***
Lower ETR couple: new job x TS					0.813**
Lower ETR couple: additional hours x TS					0.298
Lower ETR couple: additional hours x TS					‑0.087
Spending on health services x TS						‑3.131**
Spending on health services x TS						‑1.856
Other social spending in kind x TS	‑2.116*	‑2.224*	‑2.11	‑2.59**	‑3.202***	‑2.093	‑2.365*
Other social spending in kind x TS	0.16	0.617	0.343	0.774	0.848	‑0.88	0.542
Tax breaks for families x TS							‑0.211
Tax breaks for families x TS							0.789
Volume of redistribution x TS	0.296	0.215	0.434	‑0.149	0.419	0.949	0.251
Volume of redistribution x TS	‑1.139	‑1.512*	‑1.432*	‑0.829	‑1.061	‑1.615**	‑1.159
Share of taxes paid by the bottom 30% x TS	‑0.056	0.11	‑0.127	‑0.034	0.416	‑0.335	‑0.111
Share of taxes paid by the bottom 30% x TS	‑0.569	‑0.665	‑0.675*	‑0.489	‑0.973	‑0.587**	‑0.597
Share of benefits received by the bottom 30% x TS	‑1.707	‑2.414	‑1.537	‑2.197	‑0.969	‑0.851	‑1.787
Share of benefits received by the bottom 30% x TS	3.43***	3.811***	3.129**	3.474***	2.857**	4.117***	3.209***
In-cash spending on families x TS	0.834	‑0.022	0.91	1.332	0.075	0.727	0.845
In-cash spending on families x TS	3.258***	3.495***	3.21***	3.055**	2.527**	3.715***	3.335***
Total social expenditure per capita (logged)	‑8.404***	‑9.092***	‑7.923***	‑7.677***	‑8.759***	‑10.021***	‑8.724***
Adolescent fertility rate (per 1 000)	0.052	‑0.02	0.037	0.045	‑0.016	0.046	0.084
Share of single‑parent families	‑0.24	‑0.268	‑0.165	‑0.302	‑0.015	‑0.393	‑0.248
Average number of children per household	6.902	6.264	7.208	6.378	12.31	10.793	0.337
Mean poverty gap	0.536***	0.469***	0.497***	0.522***	0.549***	0.45***	0.531***
Adjusted R²	0.673	0.664	0.677	0.695	0.695	0.697	0.675
Number of observations	448	448	448	448	431	448	448
Number of countries	27	27	27	27	27	27	27
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes

	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	‑1.329	‑1.556	‑1.442	‑1.403	‑1.316	‑2.342***	‑2.268**	‑1.862**	‑1.951**
Spending on ALMP (lagged 3 years) x TS (lagged 3 years)	0.699	1.045	1.04	1.051	1.013	1.283*	1.458***	1.135**	1.289**
ECEC enrolment 0‑2 x TS	‑0.86	‑0.664	‑1.739	‑1.269	‑1.565	‑3.165**	‑2.715**	‑2.83***	‑2.529***
ECEC enrolment 0‑2 x TS	0.219	0.036	‑0.517	‑0.536	‑0.343	0.66	0.234	0.46	0.353
ECEC enrolment 3‑5 x TS	‑1.093	‑0.299	‑0.146	‑0.56	‑0.651	0.032	‑1.382**	‑1.263*	‑1.094
ECEC enrolment 3‑5 x TS	2.216***	2.557***	1.926**	2.269**	2.463***	1.16	1.64**	2.009***	2.001***
Other social spending in kind x TS	‑3.073**	‑2.565**	‑2.806***	‑2.22**	‑1.742	‑4.971***	‑1.408	0.1	0.042
Other social spending in kind x TS	1.882**	1.558	0.497	0.16	‑0.339	3.353***	0.493	‑0.86	‑0.817
Volume of redistribution x TS			0.986	0.512	0.225		‑0.471	‑0.689	‑0.578
Volume of redistribution x TS			‑1.221	‑1.083	‑1.306		0.01	‑0.098	‑0.084
Size of benefits	3.393***	3.812***
Size of taxes	0.611	‑0.276
Share of taxes paid by the bottom 30% x TS	‑0.218		‑0.079	‑0.095	‑0.018		1.209	0.883	0.926
Share of taxes paid by the bottom 30% x TS			‑1.112	‑1.233	‑1.082		‑0.754	‑0.874	‑0.924
Share of benefits received by the bottom 30% x TS	1.733		‑1.578	‑2.115	‑2.277		‑1.579	‑2.243	‑2.36
Share of benefits received by the bottom 30% x TS			3.766***	4.024***	4.072***		2.464***	2.902***	3.069***
Size of benefits x share of benefits received by the bottom 30%	‑0.669
Size of taxes x share of taxes received by the bottom 30%	‑0.823***
Income transfers for 1 child x TS			2.184***
Income transfers for 1 child x TS			‑1.939***
Income transfers for 2 children x TS				1.752**
Income transfers for 2 children x TS				‑1.31**
Income transfer for 3 children x TS					2.302***
Income transfer for 3 children x TS					‑0.889
RED index x TS						3.757***
RED index x TS						‑1.595***
In-cash spending on families x TS	‑0.695	‑0.648				‑1.723*
In-cash spending on families x TS	2.933***	2.957***				2.4**
In-cash spending on families, ages 0‑5 x TS							‑0.329
In-cash spending on families, ages 0‑5 x TS							0.285
In-cash spending on families, ages 6‑11 x TS								1.737***
In-cash spending on families, ages 6‑11 x TS								‑0.817
In-cash spending on families, ages 12‑17 x TS									2.034***
In-cash spending on families, ages 12‑17 x TS									‑0.856
Total social expenditure per capita (logged)	‑9.529***	‑9.875***	‑10.172***	‑9.887***	‑11.179***	‑12.798***	‑11.123***	‑11.358***	‑11.002***
Adolescent fertility rate (per 1 000)	0.2	0.237	0.056	‑0.002	0.026	0.21	‑0.184	‑0.191	‑0.169
Share of single‑parent families	‑0.298	‑0.223	‑0.29	‑0.107	0.015	‑0.415	‑0.152	‑0.21	‑0.238
Average number of children per household	‑2.832	‑2.577	36.764*	35.446*	33.643*	9.915	‑11.364	‑13.526	‑14.276
Mean poverty gap	0.173	0.205*	0.478***	0.519***	0.484***	0.457***	0.475***	0.476***	0.472***
Adjusted R²	0.678	0.651	0.676	0.66	0.672	0.674	0.711	0.722	0.724
Number of observations	448	448	438	438	438	438	322	322	322
Number of countries	27	27	27	27	27	27	24	24	24
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Spending on ALMP (lagged 3 years) x GDP (lagged 3 years)	‑0.035	‑0.034	‑0.033	‑0.034	‑0.039*	‑0.043**	‑0.024
Spending on ALMP (lagged 3 years) x GDP (lagged 3 years)	‑0.046*	‑0.051**	‑0.041**	‑0.052**	‑0.027	‑0.04**	‑0.055**
ECEC enrolment 0‑2 x GDP	0.047		0.03	0.048	0.053	0.046	0.052
ECEC enrolment 0‑2 x GDP	‑0.047***		‑0.038***	‑0.043**	‑0.035**	‑0.043***	‑0.046***
ECEC enrolment 3‑5 x GDP	‑0.022		‑0.009	‑0.025	‑0.018	‑0.015	‑0.014
ECEC enrolment 3‑5 x GDP	0.02		0.017	0.019	0.025*	0.015	0.009
Spending on ECEC x GDP		‑0.022	‑0.022
Spending on ECEC x GDP		0.016	0.007
ECEC enrolment 0‑2 x spending on ECEC			0.007
ECEC enrolment 3‑5 x spending on ECEC			0.007
Lower ETR single: new job x GDP				‑0.007
Lower ETR single: new job x GDP				‑0.026
Lower ETR single: additional hours x GDP				‑0.003
Lower ETR single: additional hours x GDP				0.003
Lower ETR couple: new job x GDP					0
Lower ETR couple: new job x GDP					‑0.023**
Lower ETR couple: additional hours x GDP					0.007
Lower ETR couple: additional hours x GDP					0.027**
Spending on health services x GDP						‑0.11***
Spending on health services x GDP						‑0.014
Other social spending in kind x GDP	0.058**	0.064**	0.06**	0.062***	0.068***	0.03	0.036
Other social spending in kind x GDP	‑0.004	‑0.026	‑0.016	‑0.015	‑0.024	‑0.015	0.018
Tax breaks for families x GDP							‑0.03**
Tax breaks for families x GDP							0.028***
Volume of redistribution x GDP	0.029	0.014	0.015	0.031	0.026	0.045*	0.027
Volume of redistribution x GDP	‑0.046***	‑0.029	‑0.032*	‑0.044***	‑0.06***	‑0.065***	‑0.042***
Share of taxes paid by the bottom 30% x GDP	‑0.029	‑0.026	‑0.023	‑0.032	‑0.019	‑0.018	‑0.038**
Share of taxes paid by the bottom 30% x GDP	0.002	0.015	0.015	0.001	0	0.008	‑0.003
Share of benefits received by the bottom 30% x GDP	0.001	0.006	0.004	‑0.011	‑0.006	0.017	0.007
Share of benefits received by the bottom 30% x GDP	‑0.051***	‑0.036**	‑0.033*	‑0.043**	‑0.041**	‑0.056**	‑0.061***
In-cash spending on families x GDP	0.062***	0.057***	0.062***	0.064***	0.052**	0.033	0.056**
In-cash spending on families x GDP	‑0.002	0.017	0.006	‑0.004	‑0.013	0.007	‑0.007
Ln (GDP p.c.)	0.178***	0.178***	0.172***	0.195***	0.17***	0.229***	0.179***
Ln (life expectancy)	0.38	0.238	‑0.111	0.677	1.443	‑0.948	1.087
Dependency ratio	‑0.008	‑0.006	‑0.008	‑0.006	‑0.007	‑0.003	‑0.007
Adjusted R²	0.619	0.636	0.652	0.623	0.572	0.658	0.644
Number of observations	482	482	482	482	465	482	482
Number of countries	32	32	32	32	32	32	32
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Note: Coefficients in bold are the ones used for the value for money calculations. “x” signifies an interaction between two variables. “x GDP” means the interaction between the policy variable in the line above and the log of GDP per capita. The log of GDP per capita is standardised along with the policy variables to facilitate the interpretation of results. The dependent variable is the log of total social spending per capita.

	(8)	(9)	(10)	(11)	(12)	(13)	(14)	(15)	(16)
Spending on ALMP (lagged 3 years) x GDP (lagged 3 years)	‑0.01	‑0.017	‑0.039*	‑0.033	‑0.032	‑0.041*	‑0.017	‑0.015	‑0.016
Spending on ALMP (lagged 3 years) x GDP (lagged 3 years)	‑0.045**	‑0.044**	‑0.009	‑0.008	‑0.015	‑0.03	‑0.022	‑0.026	‑0.023
ECEC enrolment 0‑2 x GDP	0.041	0.059*	0.024	0.025	0.02	0.022	‑0.001	‑0.004	‑0.001
ECEC enrolment 0‑2 x GDP	‑0.043***	‑0.045***	‑0.038**	‑0.038**	‑0.036**	‑0.031	‑0.02	‑0.027**	‑0.03**
ECEC enrolment 3‑5 x GDP	‑0.054***	‑0.046**	‑0.001	‑0.018	‑0.029	‑0.024	‑0.02	‑0.029	‑0.026
ECEC enrolment 3‑5 x GDP	0.001	‑0.001	‑0.002	0.007	0.012	‑0.005	0.015	0.016	0.016
Other social spending in kind x GDP	0.016	0.033	0.056**	0.062***	0.072***	0.055**	0.062***	0.067***	0.069***
Other social spending in kind x GDP	0.008	‑0.003	‑0.026	‑0.031	‑0.031	‑0.006	‑0.083***	‑0.083***	‑0.085***
Volume of redistribution x GDP			0.032	0.032	0.021		0.061***	0.056***	0.056***
Volume of redistribution x GDP			‑0.058***	‑0.053***	‑0.052***		‑0.067***	‑0.056***	‑0.059***
Size of benefits	0.133***	0.105***
Size of taxes	‑0.03	‑0.048**
Share of taxes paid by the bottom 30% x GDP	‑0.027		‑0.01	‑0.018	‑0.017		0.005	0.003	0.003
Share of taxes paid by the bottom 30% x GDP			0.004	‑0.002	‑0.001		0.029**	0.027*	0.026*
Share of benefits received by the bottom 30% x GDP	0.059		0.014	0.006	0.002		0.086*	0.087*	0.082*
Share of benefits received by the bottom 30% x GDP			‑0.035*	‑0.033	‑0.037*		‑0.047***	‑0.038**	‑0.037**
Size of benefits x share of benefits received by the bottom 30%	0.026**
Size of taxes x share of taxes received by the bottom 30%	‑0.014
Income transfers for 1 child x GDP			0.046***
Income transfers for 1 child x GDP			‑0.003
Income transfers for 2 children x GDP				0.05***
Income transfers for 2 children x GDP				0.006
Income transfer for 3 children x GDP					0.062***
Income transfer for 3 children x GDP					0.001
RED index x GDP						0.057**
RED index x GDP						‑0.012
In-cash spending on families x GDP	0.03	0.031				0.025
In-cash spending on families x GDP	‑0.013	‑0.018				‑0.015
In-cash spending on families, ages 0‑5 x GDP							0.014
In-cash spending on families, ages 0‑5 x GDP							‑0.001
In-cash spending on families, ages 6‑11 x GDP								‑0.003
In-cash spending on families, ages 6‑11 x GDP								‑0.032
In-cash spending on families, ages 12‑17 x GDP									0.004
In-cash spending on families, ages 12‑17 x GDP									‑0.032
GDP per capita (logged)	0.274***	0.264***	0.157***	0.161***	0.169***	0.168***	0.08	0.091*	0.094*
Life expectancy (logged)	0.604	0.875	0.454	0.527	0.024	0.079	2.278	2.222	2.219
Dependency ratio	‑0.003	‑0.002	‑0.006	‑0.007	‑0.008	‑0.008	‑0.006	‑0.007	‑0.007
Adjusted R²	0.675	0.657	0.568	0.567	0.587	0.553	0.616	0.632	0.627
Number of observations	478	478	469	469	469	469	333	333	333
Number of countries	32	32	32	32	32	32	28	28	28
Country fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Time fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Leads and lags for core and policy variables	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Annex 3.E. Value for money at different levels of social expenditure and GDP

Copy link to Annex 3.E. Value for money at different levels of social expenditure and GDP

Annex Figure 3.E.1. The value for money for a hypothetical country with relatively low levels of per capita social expenditure and GDP per capita
Copy link to Annex Figure 3.E.1. The value for money for a hypothetical country with relatively low levels of per capita social expenditure and GDP per capita

Annex Figure 3.E.2. The value for money of social policies for a hypothetical country with relatively high levels of per capita social expenditure and GDP per capita
Copy link to Annex Figure 3.E.2. The value for money of social policies for a hypothetical country with relatively high levels of per capita social expenditure and GDP per capita

Annex 3.F. Indexation of major child cash benefit amounts to inflation or cost-of-living in selected OECD countries

Copy link to Annex 3.F. Indexation of major child cash benefit amounts to inflation or cost-of-living in selected OECD countries

Country	Child/Family Benefit Indexation to Inflation? (Recent Practice)	Notes / References
Belgium	Yes (regionally automatic)	In Brussels and Wallonia, family/child benefit amounts are indexed to the health index (a cost-of-living measure) and adjusted automatically when the pivot index is reached (e.g. early 2025), (OECD, 2024[167]).
Canada	Yes (annual CPI)	The Canada Child Benefit (a taxable, income‑tested cash benefit) is indexed annually to inflation each July using the CPI to maintain purchasing power (Gonzales, 2025[168]).
France	De facto (annual adjustments)	Family benefit base (BMAF) and child benefit rates are typically uprated annually in line with anticipated CPI changes (e.g. revaluation on 1 April each year), though exact indexation rules reflect policy choices rather than strict automatic indexing (Cleiss, 2026[169]).
Austria	No automatic indexation	Austrian family allowance (Familienbeihilfe) is not indexed to inflation under current policy settings (OECD, 2024[170]).
Luxembourg	Yes (price/wage linked)	Family allowances have been indexed again since October 2021; they are linked like the minimum wage indexation mechanism (e.g. indexed to inflation/wage movements) (OECD, 2014[171]).
United Kingdom	Historically yes but conditional	UK child benefit has historically been updated annually, often to keep pace with inflation or the “highest of prices, wages or 2.5 %” for certain benefits; indexation practice depends on fiscal policy decisions (OECD, 2025[172]).
Germany	No clear automatic CPI indexing	Child benefit reforms have sometimes involved ad hoc increases responding to inflationary periods (e.g. one‑off inflation adjustments in 2022), but there’s no strict CPI indexation rule in the standard benefit description (OECD, 2023[126]).
Spain/ Italy/ other EU	Varies, often discretionary	Many EU countries (e.g. Spain, Italy) do not have automatic inflation indexation for child benefits; increases are typically set by budgetary or legislative decision rather than strict indexation. Data on rates and expenditure trends are available, but specific automatic CPI linkage varies by country (Eurostat, n.d.[173]).

Notes

Copy link to Notes

← 1. Indicators of financial strain – whether self-reported or objective (e.g. the ability to cover three months of expenses using available assets) – can help provide a more comprehensive picture of households’ economic pressure and experienced stress. However, the lack of consistent OECD-wide time series data limits their use in analysis such as in the present chapter. Similarly, combined measures such as AROPE are not available across the OECD and limited to recent years, and therefore not suitable for this analysis.

← 2. To account for differences in household size and composition, income is “equivalised,” meaning that total household income is divided by the square root of the household size. This adjustment reflects variations in living standards across families of different sizes.

← 3. In most countries, average living standards have increased over time, causing the relative poverty line – defined as a share of median income – to rise accordingly. As a result, relative poverty may include households whose living standards would not have been considered poor in earlier periods. In such contexts, anchored child poverty captures children in deeper poverty, that is, those with equivalised incomes below both the fixed (anchored) and the contemporaneous relative poverty thresholds; anchored poverty rates are therefore typically lower than relative poverty rates. However, anchored poverty rates can exceed relative poverty rates when median living standards decline, signalling a deterioration in living conditions that also affects children at the bottom of the income distribution.

← 4. This indicator was prioritised over the child-specific material deprivation rate and the material and social deprivation rate of children because it offers longer annual time series that can be consistently matched with indicators of social expenditure and other policy design.

← 5. Spending shares are not adjusted for the size of potential beneficiary groups because the analysis aims to assess the effectiveness of overall spending allocations across policy areas and the trade‑offs they imply for policymakers. Population-weighted robustness checks (e.g. weighting ECEC spending by the population under age 5 and family benefits by the population under age 20) produce broadly similar results, confirming the robustness of the main findings, although health spending (weighted by the share of the population under age 65) becomes more strongly associated with lower severe material deprivation at higher levels of social spending (results available upon request).

← 6. The term “taxes” is used in a very wide sense here and refers to reductions in income through both, taxes and social contributions.

← 7. The volume of redistribution is measured by equivalised net benefits received from public transfers (benefits net of taxes paid) by the average working-age individual, expressed as a percentage of GDP per capita. The share of benefits received by the bottom 30% measures public social security cash benefits received by working-age individuals in the bottom 30% of the post-transfer, equivalised household income distribution, per equivalent household member, as a share of benefits received by all working-age individuals. The share of taxes paid by the bottom 30% is defined analogously, as taxes paid by working-age individuals in the bottom 30% of the post-transfer, equivalised household income distribution, per equivalent household member, as a share of taxes paid by all working-age individuals. These two indicators capture the distribution of benefits and taxes across the income distribution. The share of cash transfers to families measures cash expenditure on family benefits as a share of total social expenditure. Three additional indicators measure the share of total social expenditure provided in cash to children and their families during early childhood (ages 0‑5), middle childhood (ages 6‑11) and adolescence (ages 12‑17). Tax breaks for families are measured as the amount of tax revenue forgone through tax provisions for families with children, expressed as a percentage of GDP and adjusted for the population share aged under 20. These provisions include child benefits excluded from the tax base, child tax allowances and child tax credits; where tax credits are paid out in cash, they are recorded as cash transfers. The RED index captures the joint redistributive effect of taxes and benefits, combining their progressivity and size. It is based on taxes and benefits calculated for hypothetical households covering single‑parent and couple families with different numbers of children and work intensities at low to middle earnings. A value of zero corresponds to a schedule that is exactly proportional to gross income, while higher values indicate greater redistribution towards lower-income families. Finally, income transfers for one, two and three children measure the additional tax and benefit advantages received by hypothetical families with children relative to otherwise comparable households without children, providing measures of horizontal redistribution towards families with children. For the benefit and tax distribution indicators, households are ranked according to their post-transfer equivalised income because of data availability. This may underestimate the redistributive effect of benefits or taxes where these themselves cause individuals to move into or out of the bottom 30% of the distribution.

← 8. This is based on the average sample growth in the logged values of per capita social expenditure and GDP per capita in PPP constant prices from 1997 to 2021.

← 9. The results show estimates that include country and time fixed effects, ensuring that the discussion is grounded in within-country associations between the two dependent variables and their explanatory factors.

← 10. See Annex Table 3.D.1‑6 (specification 1).

← 11. See Annex Tables 3.D.7 and 8.

← 12. Evidence of this pattern is reflected in the estimation results. When no interaction with total social spending is included, within-country increases in ECEC coverage for children aged 3‑5 are associated, on average, with a negative (though statistically insignificant) change in deprivation. Once an interaction with total social spending is introduced, the main coefficient – capturing changes in coverage around the country average – remains negative (and not statistically significant), while the interaction term becomes positive and significant. This indicates that, in higher-spending welfare systems, further increases in preschool coverage relative to a country’s average are associated with a lower efficacy at preventing severe material deprivation among children compared to the average social policy mix.

← 13. An alternative specification scales the share of social expenditures allocated to ALMP by the share of unemployed individuals in the working-age population to approximate exposure among likely beneficiaries. While intuitively appealing, this changes both the interpretation and statistical properties of the variable, shifting it from a measure of overall policy orientation towards a narrower measure of relative spending intensity for the unemployed. The results for this alternative measure become largely insignificant (available upon request), which can be expected for several reasons: As many ALMPs also target inactive and employed individuals, this scaling may understate their broader reach and thus lead to an attenuation of the estimated effects. In addition, incorporating unemployment introduces a cyclical and potentially endogenous component linked to both child poverty and social spending, contributing to smaller and less precisely estimated coefficients. For these reasons, the baseline specification using the unweighted ALMP spending share is preferred, although the sensitivity of results suggests that effects may differ depending on the populations ALMPs are intended to support and the size of the unemployed population.

← 14. Higher take‑home pay is sometimes associated with a reduced effectiveness of social expenditure to mitigate child relative poverty. However, caution is warranted when interpreting results based on relative income poverty measures, which adjust annually to changes in the median income. Broad reductions in ETRs across lower income ranges may raise median equivalised income, which can mechanically increase the share of children falling below the relative poverty threshold even when the disposable incomes of low-income families improve. The child anchored poverty rate is not impacted in this way as it relates to the poverty threshold in 2005 and estimates of severe material deprivation are unrelated to median income.

← 15. The value for money approach is slightly adjusted for spending on ECEC as a share of total social expenditure in Figure 3.10, which is found to have no significant impact on any of the child poverty measures by itself. Instead, it appears that the impact on reducing income poverty is driven by higher enrolment, likely because it enables parents to raise their labour market participation and hence their earnings (see Section 3.4). To reflect this, the value for money calculation in Figure 3.10 assumes that the simulated one‑p.p. increase in the share of social expenditure allocated to ECEC services is invested entirely in expanding enrolment, namely to finance an expansion of pre‑primary education and care (ages 3‑5) by ten per cent, and a rise in enrolment at ages 0‑2 with the remainder of the budget (+18 p.p.). The enrolment increases corresponding to a one‑p.p. increase in the share of social spending dedicated to ECEC are estimated based on the average spending-to‑enrolment ratios across countries and years in the sample. An enrolment increase of ten p.p. for children aged 3‑5 represents near full enrolment on average across the OECD in 2023. The regression coefficients underlying the value for money calculations for “spending on ECEC to increase enrolment” are taken from specification 3 in Annex 3.B.

← 16. The use of aggregate spending data in this analysis limits the ability to explore these mechanisms in detail, and the existing causal evidence on the translation of employment support into improved child material living conditions remains limited (Box 3.6).

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment

Governance

Health

Industry, business and entrepreneurship

Regional, rural and urban development

Science, technology and innovation

Society

Taxation

Trade

Energy

Nuclear energy

Transport

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Countries

Countries A - C

Countries D - I

Countries J - M

Countries N - R

Countries S - T

Countries U - Z

Regional and global engagement

Publications

Publications

Featured publications

Data

Data

Featured data

News & events

News & events

Featured events

About OECD

About

Engage with us

Work with us

Featured topics

Agriculture and fisheries

Climate change

Development

Digital

Economy

Education and skills

Employment

Environment

Finance and investment