Reducing Inequalities by Investing in Early Childhood Education and Care

With rising income inequalities and poverty within many OECD countries (see Chapter 3), rising immigration and refugee crisis (see Chapter 7), increasing attention is put on early childhood education and care (ECEC) policies as a way to mitigate long-term inequalities and support inclusion, together with other policies later in life. High-quality ECEC available to all children (see Chapters 5, 6 and 7) can provide opportunities for learning and development through rich interactions with ECEC staff and peers. By investing in these policies, countries aim to level the playing field from the early years, get lasting effects throughout childhood and adulthood, and thereby mitigate inequalities in the short-, medium- and long-term. If ECEC policies have long-lasting effects on inequalities, this provides a strong argument for investing in ECEC, as the benefits apply to a long period of time.

The chapter discusses the mechanisms behind long-lasting effects of ECEC policies on inequalities. More specifically, this chapter addresses the following questions:

• To what extent is there some evidence of long-lasting positive effects of ECEC policies on inequalities?

• What are the factors that play a role in the scale of the effects of investment in ECEC, avoiding fading out of effects and leading to long-term mitigation of inequalities? What are the policy implications?

For ECEC to mitigate inequalities in the long run, these policies need to have influence on individual outcomes later in life such as performance in higher levels of education, skills development over life, or various other adulthood outcomes (e.g. health and labour market performance), but also, stronger effects for vulnerable children than for those from advantaged socio-economic and cultural backgrounds. The chapter begins by presenting a conceptual framework that explains what is meant under the concept of reaching long-lasting effects on inequalities. It then discusses some of the main research studies that have estimated the long-lasting effects of ECEC policies on inequalities. Finally, it discusses the main features that explain these effects, and draws broad policy implications for the design of ECEC policies that are discussed more in-depth in other chapters of this report.

This chapter mostly discusses inequalities relating to children’s socio-economic background, which are shaped by a broad range of policies. It focusses on ECEC and its connections to other closely related education and social institutions and policies related to families and children, leaving aside broader economic and social policies that play a key role on inequalities, such as taxation, labour market and housing policies.

Discussions on the effects and returns of ECEC policies have driven the economic approach to ECEC and education more generally and, in many countries, have supported public investment in ECEC. This section proposes a framework to understand the long-lasting effects of ECEC policies on inequalities and explains the methodological challenges that underlie estimates of these effects.

From a theoretical perspective, there are two necessary conditions for ECEC to mitigate inequalities in the medium- to long-term:

• Condition 1: ECEC has positive long-lasting effects on some areas of children’s development with implications for their education and labour market outcomes, as well as on their success in life more generally.

• Condition 2: Positive effects are greater for vulnerable children than for others.

Studies have yielded different estimates of the short- to long-term effects of ECEC on inequalities that are discussed later in this chapter. To understand them as well as what is meant by “lasting effects on inequalities”, Figure 8.1 proposes a theoretical framework with various hypothetical scenarios of the effects of ECEC on inequalities. In a first hypothetical scenario (A) in Figure 8.1, ECEC increases outcomes for children from both low and high socio-economic backgrounds with the same magnitude. In this scenario, ECEC has no effect on the gap in performance between children from low and high socio-economic backgrounds, which remains the same throughout life. In addition, while the effects for both groups of children remain positive throughout life, they decrease over time. This scenario could, for instance, correspond to universal ECEC leading to similar positive effects for all children. In that case, Condition 1 is met but Condition 2 is not met and ECEC therefore does not mitigate inequalities. An unchanged gap in inequalities is considered as the counterfactual in the other scenarios.

In Scenario B of Figure 8.1, ECEC has positive effects on children’s outcomes from both low and high socio-economic backgrounds, but the effect is stronger compared to the counterfactual for children from low socio-economic background in the short, medium and long terms. Effects are diminishing with time but remain. In this scenario, ECEC mitigates inequalities in the short, medium and long terms. In Scenario C, effects are also positive for children from both backgrounds, stronger in the short and medium term for children from low socio-economic background, but diminishes more for those children, which leads to a situation where the long-term gap is equal to the initial gap. In this scenario, ECEC mitigates inequalities in the short and medium terms, but not in the long-term. Finally, in Scenario D, ECEC effects are positive for children from both backgrounds, but they are stronger for children from low socio-economic background only in the medium and long terms. In this scenario, ECEC does not mitigate inequalities in the short-term, but does so in the medium and long terms.

Many other types of configurations of effects are possible, but Figure 8.1 highlights important considerations on how to measure and interpret effects of ECEC policies.

• The timing of the effects is an important policy consideration. Like for most policies, effects are likely to decline over time, although not necessarily in a monotonous way. Ideally, effects should therefore be measured at different points in time. For instance, in the latter three scenarios (B to D) of Figure 8.1, the outcome gap is decreased, but not at the same time. This is an important consideration, as some studies have found no effect of an ECEC intervention in primary school outcomes between children from various backgrounds followed by effects on academic attainment and salary in adulthood that are higher for children from low socio-economic backgrounds (United States Boston programme, discussed later on this chapter) as shown in Scenario D.

• The economic approach to decisions such as investing in different types of ECEC policies or in ECEC policies versus education policies later in life, is based on looking at the present discounted value of effects on the full life cycle, i.e. the full strike-through areas in Scenarios B to D, as indicated in Figure 8.1. Decisions should not be taken by comparing the effect at a given point in time. In Scenario C, the long-term gap equals the initial gap, but the policy might have brought benefits by reducing inequalities before age 30. Overall, this framework highlights that policies on young children can have longer-term effects than those on adults (Heckman, 2006[1]). Their returns are reaped over a longer period of time than interventions later in life, which provides a strong argument for investing in ECEC policies.

Findings on the size of the effect of ECEC depend on the counterfactual, i.e. the type of education and care the children would be exposed to without the policy. As children from low socio-economic backgrounds are more at risk of experiencing lower quality of home environment than socio-economically advantaged children (see Chapter 3), the effects of ECEC are expected to be stronger, on average, for children from low socio-economic backgrounds. In addition, the counterfactual mode of education and care for children in low-income families includes informal arrangements such as relatives, friends and neighbours, while high-income parents are more likely to find alternative modes of care or informal care of higher quality (see Chapter 5). By design, ECEC programmes targeted to vulnerable children that have positive effects are more likely than universal ones to mitigate inequalities as other children do not benefit from them, and therefore have their counterfactual unchanged. However, these programmes might not meet Condition 1 if, for instance, they lead to stigmatisation effects that penalise children’s education pathways or if more advantaged children enrol in programmes of equal or better quality.

With rising education levels among the population and the development of health and welfare services in many OECD countries, the quality of the home environment, and therefore the counterfactual, has improved in families with relatively low socio-economic status (see Annex A, Workshop 5). This means that ECEC policies have less potential to lead to positive impacts on these children (Bustamante et al., 2021[2]). As highlighted in Chapter 3, families with high socio-economic status also develop a range of strategies to support their children in multiple development areas and are better equipped to take advantage of universal programmes, including ECEC programmes that have developed in many OECD countries (see Chapter 5). Overall, the improvement of the counterfactual for low-income families, the development of universal ECEC and the capacities of socio-economically advantaged families to make the most of these programmes tend to limit the effect that ECEC can have on long-term inequalities.

Lack or limits of evidence on the long-term effects of ECEC come from the fact that many ECEC policies are not evaluated, and that identification of these effects is difficult. Another reason for difficulty in measuring effects relates to the large variety of possible effects on children and later outcomes. Scenarios presented in Figure 8.1 show effects of ECEC without specifying the types of student or adult outcomes. Studies generally look at early academic cognitive skills in primary schools and possibly broader cognitive outcomes for short-term effects and a variety of medium- to long-term effects such as high school graduation rates and a range of adult outcomes (labour market and others such as incarceration, age of marriage) (Bustamante et al., 2021[2]). Education outcomes are generally measured through cognitive and academic test scores. A smaller number of studies has focussed on social and emotional skills or adopted assessment mechanisms specific to ECEC. Overall, many outcomes are generally not captured. Another reason comes from the difficulty to identify causal effects, especially for universal ECEC as all children are enrolled.

Beyond the direct effect that ECEC policies can have on vulnerable children’s cognitive and social and emotional development, ECEC policies can mitigate inequalities through several other important channels that can be difficult to capture:

• Some interventions later in life (such as in primary schools) achieve good returns because they rely on the strong bases built by ECEC policies (see later in this chapter the discussion on “dynamic complementarities”, Heckman, 2006_[1]). This mechanism could, for instance, explain Scenario D in Figure 8.1.

• In the same vein, an important aspect of ECEC policies is that they can act against potentially irreversible effects of unfavourable environments on children’s development (see Chapter 3). The very first years therefore offer an opportunity window for effective interventions that in some cases cannot be reproduced later in life. The possibility for ECEC to reverse detrimental developmental trajectories produced by stress and other factors means that children might need lower levels of support in later years if they have benefited from quality ECEC. This effect is generally not taken into account in studies.

• ECEC supports parents’ employment and thereby increases families’ income, which leads to a variety of economic benefits (e.g. stronger consumption, higher tax revenues and lower social benefit expenditure) as well as indirect benefits to children by raising the family income and lowering the risk of poverty (see Chapter 3). There is evidence that parents' earnings gains are substantial and persist for several years after the end of ECEC (Humphries et al., 2024[3]). Improved conditions for children are therefore also expected to last.

Policy decisions are often made by looking at which policies are most cost effective. For instance, as discussed as part of the critique of the Heckman curve, there are also many other policies later in life that can mitigate inequalities in the long-term, possibly with stronger effects than ECEC ones. For instance, a policy that starts at age 18 (e.g. training programmes for low-educated youth) might have higher effect at age 30 than ECEC policies but its effect before 18 is null. Determining which policy is most cost-effective entails having information on the full range of effects, but also of costs of these policies, over time, which depends on the features of these policies and is therefore hard to establish internationally.

Finally, while long-term effects of ECEC on inequalities are important, countries might value short-term effects in themselves. In Figure 8.1, this means valuing the strike-through area in the early years, even if this area declines with time. Countries value short-term and long-term effects differently depending on their contexts and priorities. For instance, Nordic countries generally spend more on ECEC policies than other OECD countries do to support children’s well-being, independently from effects on school readiness and other impacts of ECEC policies, although they are of course paying attention to these effects. Countries with fertility rate challenges might place more value than others on children’s well-being, while countries with dynamic demography might value labour market transitions more and therefore, longer-term effects.

A large number of studies have assessed the effect of ECEC participation on children’s outcomes, and some of them have focussed on the long-term effects. However, the effect on inequalities or on the differentiated effect between children from different socio-economic backgrounds has not been systematic. This section reviews some of these studies without aiming to be exhaustive, and discusses their long-term effects on inequalities with references to the two conditions (Conditions 1 and 2) mentioned in the beginning of this chapter and to the framework presented in Figure 8.1.

Targeted programmes are common in the United States, and a large strand of the literature relates to these programmes. These include, for instance, the Perry Preschool and the Chicago Parent-Child Partnerships projects in the 1960s, as well as the Carolina Abecedarian programme in the1970s that targeted socio-economically disadvantaged children. Studies have found positive effects of these programmes until adulthood such as on educational attainment, health and earnings (see Annex A, Workshop 5). For instance, a meta-analysis of 22 high-quality experiments of several United States ECEC programmes found that participation leads to reductions in special education placement and grade retention and increases in high school graduation rates (McCoy et al., 2017[4]). These programmes therefore meet Condition 1 (ECEC has positive effects on students’ outcomes later in life) as well as Condition 2 (effects are larger for vulnerable children than for other children) as they targeted children from low socio-economic backgrounds.

While promising, these effects cannot be simply generalised to recent ECEC policies in OECD countries. These interventions specifically served low-income children and were of high intensity and long duration. For instance, the Abecedarian intervention was offered for 12 months per year across the first 5 years of children's lives. The costs of these interventions were very high, estimated at around EUR 41 000 per child in 2023 for the Perry Preschool programme and EUR 112 000 for the Abecedarian intervention, while OECD countries spent on average EUR 11 145 per child in 2020 with the highest spending in Luxembourg at EUR 22 702 (Whitaker et al., 2023[5]).

More recent studies on pre-primary programmes giving priority to larger groups of vulnerable children in the United States have led to more mixed results than first targeted interventions (see Annex A, Workshop 5) (Burchinal et al., 2024[6]). A range of positive effects of the Boston programme, assessed through a lottery-based study, have been found in the long-term, such as higher school graduation, better standardised test scores taken for college admission and college enrolment, as well as reductions in disciplinary problems in high school. The programme has therefore met Condition 1 and interestingly, no effect on academic skills were found before high school, suggesting that some effects take time to emerge, as presented in Scenario D of Figure 8.1. A limited number of outcomes were more favourable for children from low socio-economic backgrounds than for those from more advantaged socio-economic backgrounds, meaning that Condition 2 was weakly met. In contrast, a study of the Tennessee Voluntary Pre-K program (TNVPK), a large, state-funded Pre-K programme that annually enrols about one-quarter of the state’s 4-year-olds found positive effects in literacy, language and mathematics at the end of the programme, but that these gains quickly disappeared (Condition 1 unmet). Some impacts had become negative at the third year of primary education and even more in lower secondary education. The programme was considered to meet some structural features of quality such as ratios of students to teacher, class size and teachers’ education, although its quality and intensity was lower than programmes from the 1960s-70s.

Research mostly on Europe’s ECEC systems brings evidence of the effect of universal ECEC, in contrast to studies from the United States on targeted programmes. For instance, a meta-analysis of 17 longitudinal studies in nine European countries investigates whether process quality in regular ECEC gains is associated with lasting gains on language and literacy as well as mathematics (Ulferts, Wolf and Anders, 2019[7]). It found small but positive lasting effects (Condition 1 met), but since the effect held for children from various family backgrounds, the study is not conclusive on an inequality effect (no evidence for Condition 2). Another meta-analysis looked at the effects of universal ECEC on children's outcomes focussing on studies that identify causal effects (van Huizen and Plantenga, 2018[8]). Countries covered are Australia, Canada, France, Germany, Norway, Spain, the United Kingdom and the United States. The results show that ECEC quality matters critically and do not indicate that effects are fading out in the long run. Furthermore, the gains of ECEC are concentrated within children from lower socio-economic status families, also suggesting that high-quality universal ECEC can mitigate inequalities (both Conditions met, as in Scenario B of Figure 8.1).

Nordic countries have universal systems that start at an early age and therefore provide good cases to examine the effect of universal ECEC on inequalities. Some studies have found that early gaps in language skills, existing between immigrant and non-immigrant children in Denmark at age 2 or 3 and at age 2 between low- and high-SES and immigrant and non-immigrant children in Norway remain rather stable during the pre-primary period, and do not reduce before the introduction into primary school (Leseman and Slot, 2020[9]). However, a study of the expansion of subsidised childcare in Norway in the 1970s found positive adult outcomes (e.g. education, labour market attachment) in individuals’ early 30s (Havnes and Mogstad, 2011[10]). Most of the effect on education concerned children with low-educated mothers, whereas most of the effect on labour market attachment and earnings related to girls. The study concludes that subsidised ECEC increases intergenerational mobility and closes the gender-wage gap (both conditions are met). The reason could be that Nordic systems include limited explicit exposure to language education and therefore lead to effects in line with Scenario D, with small (or no) short-term effects but effects in the longer term. A study comparing intergenerational inequalities in Denmark and the United States and looking at various outcomes found that family influence on many children outcomes is comparable in both countries, despite the generosity of the Danish welfare system (Heckman and Landersø, 2022[11]). The reason is that family influence operates through various channels and might offset the mitigating effect of ECEC on inequalities. Condition 2 is therefore not met because of some compensatory effect outside the strict ECEC sphere (see later section on engaging parents and families).

In the United States, a longitudinal study of children in community-based ECEC programmes found reduced disparities between low- and higher-income children's educational attainment and wages at age 26, and that disparities in college graduation were also reduced (see Annex A, Workshop 5) (both conditions met) (Bustamante et al., 2021[2]). A study in Israel found long-lasting effects of universal ECEC that go beyond educational attainment (DeMalach and Schlosser, 2024[12]). Boys were less likely to have a juvenile criminal record and young women tended to marry later. Effects are stronger for children with parents with relatively low levels of education. The study shows that disadvantaged communities benefit from public pre-primary education even in the absence of a well-targeted education programme. Universal ECEC also has effects on advantaged children but they are generally lower, so inequalities are mitigated (both conditions met) (DeMalach and Schlosser, 2024[12]). Furthermore, large representative cohort studies in the Netherlands and the United Kingdom found substantial reduction of the gaps in language skills, at least for migrant children, as a consequence of participation in ECEC (Leseman and Slot, 2020[9]).

As research on the long-lasting effects of ECEC has developed and led to mixed results depending on the context and features of ECEC systems, there has been a growing interest in the factors that generate long-lasting effects. This literature has large potential to inform policies, but at the same time, evidence is still limited and there are several blind spots. The quality of studies varies and identifying causal effects, instead of simple correlations, is challenging and still relatively rare. This section discusses the status of evidence on features of ECEC policies that can contribute to mitigating inequalities in the long-term.

Both the quantity of participation in ECEC and its quality matter for achieving long-term effects (Dalli et al., 2011[13]). There are two main parameters for the quantity of ECEC: the starting age and the number of hours per week (the intensity). The quality of ECEC is a broad concept that includes several dimensions (see Chapter 6 and (OECD, 2021[14])). The quantity and quality of ECEC are closely related – long and intensive participation in ECEC might lead to positive long-term outcomes, especially for vulnerable children, provided that it is of high quality. This section discusses evidence on the role and relative importance of the quantity and quality of ECEC for reaching lasting effects on children and inequalities.

Evidence on how children develop and how inequality gaps build up from a very early age as well as the “skills beget skills” theory would argue for early enrolment in ECEC (see Chapter 3). Enrolment in ECEC could help change children’s trajectories if done at an early stage, such as at age 2 and below. However, findings on the impact of early start in ECEC vary across countries and are closely related to the quality of ECEC. There is evidence that enrolment in ECEC at age 2 to 3 years is beneficial for children, while evidence is more mixed for younger children.

Concerns about ECEC enrolment at an early age (especially under the age of 1) mainly relate to children’s social and emotional development and health (Melhuish et al., 2015[15]). This is particularly the case if combined with many hours of ECEC attendance (see below), with the risk that this might be “too much too early” for children. Concerning cognitive and language development, ECEC attendance in children’s first year can, but does not necessarily, have negative effects. High-quality ECEC between 1 and 3 years old tends to have relatively positive impact on language development, early numeracy and motor skills, although effects are somewhat mixed depending on countries (Carbuccia et al., 2020[16]). For instance, evidence from a large, nationally representative French birth cohort (the Étude Longitudinale Française depuis l'Enfance – ELFE) finds that attendance at age 1 has a positive impact on language skills, no impact on motor skills, and a negative impact on behaviour (Berger, Panico and Solaz, 2021[17]). Moreover, the positive impact on language skills is particularly concentrated among disadvantaged children. Overall, studies consistently find effects to be stronger for children from low-educated and low-income families, while the starting age was found to have no impact on the test scores of children from high-income families. These results suggest that early ECEC enrolment could be targeted to low-income families, but that the age of enrolment alone is not a major factor explaining the impact of ECEC.

Evidence on the effect of the number of hours (per week) of ECEC goes in the same direction as that of the starting age. Overall, there is mixed evidence of advantage for children attending full-day ECEC for the early ages while effects are more positive for older children, especially those from socio-economically disadvantaged backgrounds. Some studies have found that intensive participation (around 30 hours and more) in ECEC from an early age has positive impacts on cognitive development and language development especially for vulnerable children, provided it is of high quality (Bigras and Lemay, 2012[18]). Evidence is less conclusive for the impact on pro-social behaviour, especially for children below age 1 for high-intensity ECEC. There is evidence that long hours in ECEC for children under age 2 are linked to more risk-taking behaviours in adolescence (Dalli et al., 2011[13]). High-intensity ECEC is more positive for children from disadvantaged socio-economic backgrounds (Melhuish et al., 2015[15]). A meta-analysis of 250 estimates from 30 studies conducted between 2005 and 2017 indicates that more intensive programmes, if of high quality, produce more favourable outcomes and that the gains of ECEC are concentrated within children from lower socio-economic backgrounds (van Huizen and Plantenga, 2018[8]).

There is a consensus that ECEC needs to be of high quality to lead to positive lasting effects on children (OECD, 2021[14]). For instance, a meta-analysis of the effects of universal ECEC on children's outcomes found that ECEC quality matters critically (van Huizen and Plantenga, 2018[8]). Two indicators of quality were considered in the study, educational levels of ECEC staff and staff-to-child ratios. The evidence does not indicate that effects are fading out in the long run. The gains of ECEC are concentrated within children from lower socio-economic backgrounds, meaning that inequalities might be reduced. Similarly, while not focussing specifically on long-term effects, another literature review found that quality is a key factor that contributes to the positive effect of ECEC on children’s development in a broad range of areas, with large evidence for academic skills as well as for social and emotional development, although these are less frequently covered (Carbuccia et al., 2020[16]) (see Annex A, Workshop 2). Reviewing various studies, a Canadian meta-analysis concludes that quality, both through structural factors (staff-to-children ratios and staff qualifications) and process factors (the quality of the interactions children experience with staff members and other children in ECEC settings), plays a core role for positive effects, especially for vulnerable children (Bigras and Lemay, 2012[18]).

A policy implication is that it is important to focus on the combination of intensity and quality. While the effect of the quantity of ECEC on children’s outcomes depends on the quality of ECEC, sustaining high-quality ECEC is also closely related to the quantity of ECEC. Because of the cumulative nature of learning, more time spent in ECEC can yield better skill development, but this assumes that ECEC programmes are able to articulate and implement learning activities that differentiate practices and activities as they develop over time based on children’s mastery of skills and content (Li et al., 2020[19]). A study on the United States Head Start programme found no difference on a range of children’s outcomes between children enrolled half-days and full-days (Leow and Wen, 2016[20]). An explanation given by the authors is that quality is not sustained during the whole day, for instance with staff becoming more tired.

Quantity and quality of ECEC can be considered together through a range of policy levers. The curriculum framework and preparation of ECEC staff as well as their working conditions (e.g. time in contact with children, and time to prepare activities) need to be adapted to the length and intensity of programmes. When programmes start at an early age and cover several years, they should follow children’s development and leverage their duration to develop various skills, balancing between strengthening skills acquisition and developing new ones. When ECEC programmes are intensive and include full days, curriculum framework guidelines as well as the organisation of the time during the ECEC day can help in alternating activities for children following their rhythms. Working conditions for staff are also important to help them sustain the effort throughout the day. For these reasons, the link between the quantity of ECEC and outcomes is not linear: effects of two years of part-time ECEC are unlikely to lead to twice the effect of a year of full-time ECEC.

Features of structural and process quality of ECEC that can support equity in and through ECEC have been discussed in Chapters 6 and 7. This section focusses on features that are particularly important for achieving long-term effects on inequalities.

Areas of development targeted by ECEC relate to the length of the effect of ECEC and the risk of fading out. A potential reason for observing short-term effects and then fading out is that many of the skills that are supported through ECEC policies are likely to be developed in other levels of education or through other interventions (Duncan et al., 2023[21]) (see Annex A, Workshop 5). To avoid fading out, ECEC needs to boost skills that are foundational or fundamental and unlikely to be as much developed in the absence of ECEC participation. Some authors have qualified these skills as “trifecta” skills, those that meet three conditions: i) being malleable so that they can be changed and developed through ECEC; ii) being fundamental for later skills development and later success, as well as central for children's capacity to navigate the transition to formal schooling; and iii) unlikely to develop in the absence of the intervention.

This third condition is particularly important for ECEC to mitigate inequalities: it needs to boost some skills in vulnerable children that they are unlikely to develop at home and throughout their education pathway, and more so than other children. This is an important but also difficult to fulfil condition, as all children are exposed to learning opportunities throughout their education pathway. The combination of fundamentality and malleability is most apparent in children’s early basic literacy and mathematics skills. However, simple academic skills are likely to develop later in education in the absence of ECEC participation. The impacts of interventions that target them may fade out most quickly, as virtually all children will eventually receive this instruction. Overall, broader cognitive skills, which are usually thought of as less sensitive to instruction (e.g. vocabulary) better meet the conditions of “trifecta” skills than narrow school-readiness skills such as identifying letters and numbers.

A study of various ECEC programmes in the United States found that impacts on cognitive measures tended to be smaller at the end of ECEC but persist longer than impacts on achievement measures (Li et al., 2020[19]). This result can be explained by the fact that children receive more and more direct instruction on achievement skills once they enter primary school, and therefore the effect of ECEC on these skills diminishes. Similarly, a potential reason for United States ECEC programmes becoming less efficient over time is that there has been a shift from using hands-on learning and focussing on promoting health, language and social skills (Abecedarian and Perry programmes) to using didactic large group instruction to teach early reading and mathematics skills (Whitaker et al., 2023[5]). For instance, the Tennessee Voluntary Pre-K programme has been found to have negative impacts on reading, mathematics, and science scores at the end of third grade for children who have been assigned to these programmes, meaning that children who attended the programme were harmed by the experience in terms of their academic skills in elementary school. The reasons for this are not clear, but one speculation is that the programme included practices and learning content similar to those of the first year of primary school.

Another type of distinction in the same vein is between “constrained” and “unconstrained” skills and ability to perform “closed” and “open” tasks (Bailey et al., 2016[22]). Constrained and closed skills require only a limited amount of knowledge and are simple enough for virtually all individuals who practice them to master. Fostering these skills early on leads to no long-term effect because children would have acquired them in any case. Many of the early academic skills fall into the “closed” category (e.g. name writing, alphabet knowledge). In contrast, ability to perform open tasks, such as general mathematics achievement or vocabulary, is always incomplete so that even extensive practice still leaves room for improvement. More sophisticated skills develop at different speeds depending on children’s home environment, with the slowest growth occurring for the most complex skills in conditions faced by vulnerable children. Thus, efforts to target more sophisticated skills for vulnerable children are more likely to mitigate inequalities in the long-term.

Following the same logic, ECEC could be thought of as launching vulnerable children on more positive “trajectories”, which has been described as bringing a “foot-in-the-door” advantage, by providing some sort of permanent increase in key skills or capacities that offers a lifetime of benefits. It is also somewhat difficult to determine which skills would lead to this outcome. However, the skills-building hypothesis suggests also having curriculum frameworks that do not focus on simple early academic skills, but include more complex early cognitive skills and communication skills, with particular efforts to develop these skills among vulnerable children. Social and emotional skills are areas of development that are important for school readiness and learning but the effect of ECEC on these skills has not been widely documented (Carbuccia et al., 2020[16]). Among these skills, self-regulatory skills have been widely studied.

Finally, rather than focussing on trifecta skills meeting the three conditions, a different approach consists in targeting important but difficult-to-change skills or behaviours with intensive interventions for subgroups of children most in need of help and least likely to develop those skills in the absence of the intervention (Bailey et al., 2016[22]). For instance, in the United States, the Abecedarian programme appears to have successfully boosted the IQ levels of children with low initial IQ scores who are living in families with multiple disadvantages. Combining intensive ECEC and a focus on less malleable skills such as general intelligence and conscientiousness is another policy direction.

Another source of positive long-term effects of broad participation in ECEC, as opposed to targeted ones, comes from the improved class climate in later years of education and possibilities for teachers in primary education to implement a more demanding curriculum (see below). Reaching these positive peer effects would suggest to target academic and social and emotional skills because they would support higher-level instructional content in subsequent years of education (Bailey et al., 2016[22]). In the same vein, a study found that the fade-out effect in education is linked to the share of classroom peers in primary education assigned to pre-primary education – with enough children having attended the same ECEC programme, social interactions among peers are stronger and the fade-out effect is mitigated (List and Uchida, 2024[23]). These positive effects are more likely to benefit children from disadvantaged socio-economic backgrounds who are often clustered into the same schools because of housing segregation.

Since parents influence their child’s development, learning and well-being and overall success in education and life more generally (see Chapter 3), supporting rich parenting behaviours in families of low socio-economic status and minority families, and thereby raising the quality of the home environment that children experience, can be a powerful tool to mitigate inequalities. Comparison of intergenerational social mobility in Denmark and the United States shows that family influence operates through multiple channels – direct parental interactions with children in stimulating learning, choices of neighbourhoods and localities, which influence the quality of schooling and the quality of peers, and guidance on important lifetime decisions (Heckman and Landersø, 2022[11]). These effects act throughout childhood and adulthood and operate even in the presence of universal benefits, as is the case in Denmark.

Since both ECEC and parenting programmes aim to support children’s development, combining them could be a promising direction to mitigate inequalities (Duncan et al., 2023[21]). Combining ECEC and parenting programmes can take various forms, such as ECEC centres informing and counselling parents through posters about appropriate practices with children, or offering possibilities for parents to talk to staff to the organisation of visits to families through ECEC centres (see Chapter 10).

Evidence on the effects of combined ECEC and household interventions on children’s skills is limited. A meta-analysis of United States ECEC services including home interventions in the 1960-70s found that not all parenting education programmes are effective at improving children's cognitive and pre-academic outcomes (Grindal et al., 2016[24]). While the study is restricted to relatively old United States interventions and looks only at short-term effects on early cognitive academic skills, it brings findings on the preferred design of parental interventions. Low quality and frequency of the parental engagement component might explain the lack of effect. Most parenting education was provided through one or two home visits a year, often focused on general topics that parents identified as being of interest. These programmes may need to offer greater frequency to produce meaningful changes in parents’ behaviour that are sustained in daily home interactions. The study found that when parenting education was provided through one or more home visits a month, the effect sizes for cognitive outcomes were significantly larger than for programmes that provided lower dosages of home visits. In addition, programmes that engage parents in active learning, through opportunities to observe and practice particular parenting skills, might have greater impacts than those that do not attempt to change parental behaviour in such a systematic way. Similarly, a study on the United States Head Start programme that consists of ECEC provision in addition to a range of medical and nutrition services, and helps parents to foster their child’s development, found that centres offering more frequent home visiting (more than three home visits per year) were effective at improving children’s behaviours and attention (Walters, 2015[25]).

Limits of programmes aiming to change parental behaviour mostly relate to the challenges of enrolling parents and keeping them engaged when programmes are of high intensity (Duncan et al., 2023[21]). On the other end, low-intensity programmes are less likely to deliver effects. In addition, evidence suggests that low socio-economic status parents tend to spend less time on rich interactions with their children because of higher level of stress and prioritisation of activities involving financial gains and time that provide immediate returns rather than long-term ones, with these two aspects being interrelated (see Chapter 3). As a result, information campaigns alone might not be very effective in changing parental behaviours. Policies that focus on reducing stress or offer less time-consuming practices that integrate well into parents’ habits are more likely to be sustained and have long-lasting effects (see Chapters 4 and 10).

The landscape of policies that influence children’s development, learning and well-being is broad (see Chapter 4). The co-ordination of ECEC policies with other levels of education and other areas that affect children’s development is crucial to achieve long-term effects.

For ECEC participation to have long effects on children, there should be continuity of high-quality experience across different stages of development of children and diverse education settings, which includes a range of features: alignment of learning expectations and curricula and co-ordination to avoid redundant content, pedagogies adapted to children’s age and continued rich interactions between teachers and children throughout education experiences, continuity in assessment and learning environments, and co-operation between staff, while co-location can also help (OECD, 2017[26]). Transitions within ECEC (e.g. from under age 3 ECEC settings to pre-primary education) and from ECEC to primary education are sensitive periods for children that can lead to a fade-out of what has been gained through their past experiences.

The focus is generally put on how the lower level of education can prepare children for the higher ones. However, a higher level of education can be designed to ensure the children do not lose and can build on what they have learnt. For instance, focussing on transitions between ECEC and primary education, the research literature highlights the importance of shaping primary education to help sustain the effects of ECEC. Some authors have argued that many primary education programmes in fact sustain or amplify inequalities rather than further mitigating them, which explains why outcomes of ECEC programmes seem to fade out (Allen and Hutton, 2023[27]). The impact of ECEC could be bolstered by modified curricula for the early years of primary education that better build on the gains produced by effective ECEC (Li et al., 2020[19]). A similar recommendation applies to transitions within ECEC.

As discussed in the previous sections, policies that exclusively build on targeting some areas of development in ECEC are unlikely to mitigate inequalities in the long-term, as most of the early skills can be developed later in life. However, a sequence of skill-building interventions across both the early childhood years and the first years of primary education that build on one another and promote dynamic complementarity during a period in which skill-building is particularly rapid can be efficient (Duncan et al., 2023[21]). The concept of dynamic complementarity relates to the fact that children with stronger cognitive and non-cognitive skills will profit more from further education. The focus needs to be on skills that are central to children’s capacity to navigate the transitions within ECEC and to formal schooling, including both cognitive and social and emotional skills. In addition, children need to be continuously exposed to learning content that is rigorous, at the upper limit of their skills and understanding, and that challenges them enough to promote new, higher-level skills (Ansari et al., 2023[28]). This requires that practices and instructions are individualised to adapt to all children’s needs, lead to experiences in which children are active (rather than being didactic) and avoid situations in which vulnerable children are exposed to redundant or less ambitious content. These dynamic complementarities mean that because of the investment made in the early years, investment made in higher levels of education leads to better outcomes as ECEC sets the ground for more learning later.

Large-scale and universal ECEC programmes translate into larger percentages of children benefiting from ECEC and being better prepared for primary education. This can generate more positive peer effects and allow teachers to push their students through more advanced curricula, thereby increasing the likelihood of sustaining ECEC gains (List and Uchida, 2024[23]). This approach would benefit vulnerable children more than others, but also benefit advantaged children and thereby allow some positive peer effects to develop while limiting risks of stigmatisation.

While ECEC can provide a warm and secure environment to children and support children’s development and learning, it is unlikely to provide the conditions for healthy development by itself and therefore needs to be combined with other policies around children and families (see Chapters 4 and 10). Co-ordination of a range of services around children not only is an efficient policy to support children’s development and parents in the early years but also a policy that is most likely to make the effects of ECEC policies last (Bailey et al., 2016[22]).

A key question to effectively co-ordinate is how to combine services and welfare supports throughout childhood to lead to more lasting effects than simple ECEC participation. The existence of critical and sensitive periods of childhood in children’s development, learning and well-being means that public investment around families and children needs to accompany these patterns (see Chapter 9).

From a research perspective, how to best combine these services to reach long-term effects has not been investigated much. There are indications that care needs to be paid to avoid the redundancy of services with similar objectives for higher cost-effectiveness. For instance, a Danish study looked at the respective effects of the first Danish public pre-primary programme (by age 3) for poor children in the 1960s and a nurse home-visiting programme that was available at the same time (Rossin-Slater and Wüst, 2020[29]). Both programmes included a health component and the nurse home-visiting programme included education about parent-child interactions. The study found substitution effects between the two programmes – access to the nurse home-visiting programme reduces the positive impact of pre-primary education on the human capital index. The study suggests that when public resources are limited, it may be efficient to design programmes that specifically target populations without prior exposure to other interventions (see Chapter 9). How multiple services around families and children can be combined to give a real boost to children, especially more vulnerable ones, is discussed in Chapter 10.

[27] Allen, L. and R. Hutton (eds.) (2023), Closing the Opportunity Gap for Young Children, National Academies Press, Washington, D.C., https://doi.org/10.17226/26743.

[28] Ansari, A. et al. (2023), “The First-Grade Outcomes of Pre-K Attendees: Examining Benefits as a Function of Skill Type, Environments, and Subgroups”, American Educational Research Journal, Vol. 60/6, pp. 1139-1173, https://doi.org/10.3102/00028312231195559.

[22] Bailey, D. et al. (2016), “Persistence and Fadeout in the Impacts of Child and Adolescent Interventions”, Journal of Research on Educational Effectiveness, Vol. 10/1, pp. 7-39, https://doi.org/10.1080/19345747.2016.1232459.

[17] Berger, L., L. Panico and A. Solaz (2021), “The Impact of Center-Based Childcare Attendance on Early Child Development: Evidence From the French Elfe Cohort”, Demography, Vol. 58/2, pp. 419-450, https://doi.org/10.1215/00703370-8977274.

[18] Bigras, N. and L. Lemay (2012), Petite enfance, services de garde éducatifs et développement des enfants: État des connaissances, Presses de l’Université du Québec.

[6] Burchinal, M. et al. (2024), “Unsettled science on longer-run effects of early education”, Science, Vol. 384/6695, pp. 506-508, https://doi.org/10.1126/science.adn2141.

[2] Bustamante, A. et al. (2021), “Adult outcomes of sustained high‐quality early child care and education: Do they vary by family income?”, Child Development, Vol. 93/2, pp. 502-523, https://doi.org/10.1111/cdev.13696.

[16] Carbuccia, L. et al. (2020), Revue de littérature sur les politiques d’accompagnement au développement des capacités des jeunes enfants, Laboratoire interdisciplinaire d’évaluation des politiques publiques, Paris.

[13] Dalli, C. et al. (2011), Quality early childhood education for under-two-year-olds: What should it look like? A literature review, Ministry of Education, New Zealand.

[12] DeMalach, E. and A. Schlosser (2024), “Short- and Long-Term Effects of Universal Preschool: Evidence from the Arab Population in Israel”, Working Papers, No. 10904, CESifo , Munich, Germany.

[21] Duncan, G. et al. (2023), “Investing in early childhood development in preschool and at home”, in Handbook of the Economics of Education, Elsevier, https://doi.org/10.1016/bs.hesedu.2022.11.005.

[24] Grindal, T. et al. (2016), “The added impact of parenting education in early childhood education programs: A meta-analysis”, Children and Youth Services Review, Vol. 70, pp. 238-249, https://doi.org/10.1016/j.childyouth.2016.09.018.

[10] Havnes, T. and M. Mogstad (2011), “No Child Left Behind: Subsidized Child Care and Children’s Long-Run Outcomes”, American Economic Journal: Economic Policy, Vol. 3/2, pp. 97-129, https://doi.org/10.1257/pol.3.2.97.

[1] Heckman, J. (2006), “Skill Formation and the Economics of Investing in Disadvantaged Children”, Science, Vol. 312/5782, pp. 1900-1902, https://doi.org/10.1126/science.1128898.

[11] Heckman, J. and R. Landersø (2022), “Lessons for Americans from Denmark about inequality and social mobility”, Labour Economics, Vol. 77, p. 101999, https://doi.org/10.1016/j.labeco.2021.101999.

[3] Humphries, J. et al. (2024), Parents’ Earnings and the Returns to Universal Pre-Kindergarten, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w33038.

[20] Leow, C. and X. Wen (2016), “Is Full Day Better Than Half Day? A Propensity Score Analysis of the Association Between Head Start Program Intensity and Children’s School Performance in Kindergarten”, Early Education and Development, Vol. 28/2, pp. 224-239, https://doi.org/10.1080/10409289.2016.1208600.

[9] Leseman, P. and P. Slot (2020), “Universal versus targeted approaches to prevent early education gaps. The Netherlands as case in point, Universelle versus zielgruppenorientierte Ansätze zur Verhinderung früher Bildungsdisparitäten – Das Beispiel der Niederlande”, Zeitschrift für Erziehungswissenschaft, Vol. 23/3, pp. 485-507, https://doi.org/10.1007/s11618-020-00948-8.

[23] List, J. and H. Uchida (2024), Here Today, Gone Tomorrow? Toward an Understanding of Fade-out in Early Childhood Education Programs, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w33027.

[19] Li, W. et al. (2020), “Timing in Early Childhood Education: How Cognitive and Achievement Program Impacts Vary by Starting Age, Program Duration, and Time Since the End of the Program”, EdWorkingPaper, No. 20-201, Annenberg Institute at Brown University.

[4] McCoy, D. et al. (2017), “Impacts of Early Childhood Education on Medium- and Long-Term Educational Outcomes”, Educational Researcher, Vol. 46/8, pp. 474-487, https://doi.org/10.3102/0013189x17737739.

[15] Melhuish, E. et al. (2015), “A review of research on the effects of early childhood Education and Care (ECEC) upon child development. CARE project”, Curriculum Quality Analysis and Impact Review of European Early Childhood Education and Care (ECEC).

[14] OECD (2021), Starting Strong VI: Supporting Meaningful Interactions in Early Childhood Education and Care, Starting Strong, OECD Publishing, Paris, https://doi.org/10.1787/f47a06ae-en.

[26] OECD (2017), Starting Strong V: Transitions from Early Childhood Education and Care to Primary Education, Starting Strong, OECD Publishing, Paris, https://doi.org/10.1787/9789264276253-en.

[29] Rossin-Slater, M. and M. Wüst (2020), “What is the Added Value of Preschool for Poor Children? Long-Term and Intergenerational Impacts and Interactions with an Infant Health Intervention”, American Economic Journal: Applied Economics, Vol. 12/3, pp. 255-286, https://doi.org/10.1257/app.20180698.

[7] Ulferts, H., K. Wolf and Y. Anders (2019), “Impact of Process Quality in Early Childhood Education and Care on Academic Outcomes: Longitudinal Meta‐Analysis”, Child Development, Vol. 90/5, pp. 1474-1489, https://doi.org/10.1111/cdev.13296.

[8] van Huizen, T. and J. Plantenga (2018), “Do children benefit from universal early childhood education and care? A meta-analysis of evidence from natural experiments”, Economics of Education Review, Vol. 66, pp. 206-222, https://doi.org/10.1016/j.econedurev.2018.08.001.

[25] Walters, C. (2015), “Inputs in the Production of Early Childhood Human Capital: Evidence from Head Start”, American Economic Journal: Applied Economics, Vol. 7/4, pp. 76-102, https://doi.org/10.1257/app.20140184.

[5] Whitaker, A. et al. (2023), Why are Preschool Programs Becoming Less Effective?, http://www.edworkingpapers.com/ai23-885.