Skip to content

Child care and early education in Australia - The Longitudinal Study of Australian Children

1. Child care and early education in Australia

1.1 Introduction

The Longitudinal Study of Australian Children (LSAC) seeks to document the breadth of children’s early experiences to understand the factors impacting on health and development. The theoretical framework upon which LSAC is based is a pathways socio-ecological model of children’s health and development (Sanson et al. 2002), which acknowledges the complex interactions between children and their environments. As Figure 1 shows, infant and child development occurs within the context of the family, non-parental child care and early education (including school), and community environments. These contexts, in turn, sit within broader socioeconomic, structural, cultural and political contexts.

Figure Description

Figure 1: Socio-ecological contexts shaping children’s development

Figure 1: Socio-ecological contexts shaping children’s development

Source: Adapted from Bronfenbrenner (1979).

This report focuses primarily on children’s family and child care/early education environments, and the relationships between these spheres of influence.

Child care, either within the family network or in settings outside of the family, is an increasingly common experience for many Australian children. According to figures released by the Australian Bureau of Statistics (ABS) (2006), regular child care is used for 34 per cent of infants under the age of 1 year, 61 per cent of 1 year olds and 71 per cent of 2 and 3 year olds. As children get older, the type of care used changes. Child care for infants under 1 year is primarily informal home-based care provided by relatives (mainly grandparents) or other adults (29 per cent). Only 7 per cent of infants under 1 year attend formal care settings, such as long day care centres or family day care, but rates increase to 31 per cent for 1 year olds, to 46 per cent for 2 year olds, and 53 per cent for 3 year olds. By age 4 years, most children are attending a formal child care or early education setting, in either a long day care (28 per cent) or preschool (62 per cent) setting (ABS 2006).

Preliminary analyses of the LSAC data set indicated that for the 4 to 5 year-old cohort the type of care or early education setting children attend is influenced by where the child lives (Harrison & Ungerer 2005). Australian states and territories differ widely in their education provisions for 4 to 5 year-old children (Press & Hayes 2000). In the majority of states/territories, it is common for government schools to provide part-time preschool programs for children in the year before they enter formal schooling. However, in the two largest states, New South Wales and Victoria, school-based preschool programs are rare and families are more likely to access preschool programs that are provided by private or community-based services. All states/territories provide full-time pre-Year 1 programs for children who are 5 years old, or who will be 5 by mid-year, but name these differently (WA: pre-primary; SA: reception; Vic. Tas. Qld: preparatory; NSW, ACT: kindergarten; NT: transition). These distinctions in the availability and cost of early education across different states/territories impact families’ decisions about their use of preschool, child care and school-age early childhood programs for their children.

A further factor affecting children’s child care experiences is families’ use of multiple child care arrangements, estimated at 2 per cent for infants under 1 year, 13 per cent for 1 year olds, 19 per cent for 2 year olds and 21 per cent for 3 year olds. Again, the pattern of care varies by age. Press (2006) comments that infants are more likely to be in a mix of informal home-based care arrangements, while 2 to 3 year olds more likely to be in a combination of formal and informal care, and 4 year olds more likely to attend two or more formal settings (including preschool).

Aspects of child care and early education experiences that impact development include: the type of care/education setting attended, particularly home versus centre care, and regulated versus unregulated care; the quality of care, which is closely linked to aspects of regulation; the amount of care received, which includes weekly hours as well as duration of time or age of entry to care; and the stability of care arrangements, which arises from the number of multiple care arrangements at any one time as well as changes to care over time.1

Consideration of how these four aspects of child care/early education impact children’s development must take into account the broader context of family circumstances and choice, as well as parental care giving. Type, quality, amount and stability of child care are often correlated with other risk and protective factors within the family. For example, child care arrangements are found to be affected by a complex array of factors that differentiate families, including family finances (Wolcott & Glezer 1995), family social class and marital status (Vandell & Corasaniti 1990), levels of stress or support in the family, maternal psychological wellbeing and quality of the marital relationship (Richters & Zahn-Waxler 1988). Notably, these factors have also been identified as important variables which directly influence the quality of maternal care giving, irrespective of child care (Belsky & Isabella 1988; Bronfenbrenner & Crouter 1982; Bronfenbrenner, Moen & Garbarino 1984) and, therefore, may partly or wholly explain associations between child care and child outcomes.

This report examines the child care and early education experiences children receive and investigates the association of these experiences on their learning and social development. It also recognises the family context and family circumstances as having a significant direct influence on children’s development, but also as key determinants of children’s experience of non-parental child care—when they start, what type or types of care they receive and for how many hours each week. A further consideration is that these experiences may also be a positive influence on families’ capacity to parent their children, by providing models of care giving or social support.

[ top ]

1.2 Report overview

This report was commissioned by the Australian Government Office for Women as part of the Women’s Leadership and Development Programme and presents findings on families’ use of child care and early education for infants and 4 to 5 year-old children and the relations between care/education, family circumstances and infant/child health and development. It uses the LSAC Wave 1 data set, collected from March to November 2004, to investigate:

This report begins with a comprehensive review of international literature on non-parental child care and early education. Key themes arising from this literature inform the analyses presented in subsequent sections. The LSAC study design and methodology is outlined in Section 2. Sections 3 and 4 present the data from the infant cohort, to describe patterns of child care and the associations between care and diverse family, child and community circumstances. Sections 5 and 6 present data from the child cohort, to describe patterns of care and early education, and the associations between children’s experiences and family, child and community circumstances. Section 7 examines structural and process indicators of quality in child care and early education settings to describe variations in quality across settings. Section 8 examines infant health and development and child social and cognitive outcomes, in a series of multivariate analyses designed to assess the unique contribution of child care/early education factors after accounting for the effects of family, child and community circumstances. Section 9 summarises and highlights key findings from these descriptive and statistical analyses in relation to current issues in early childhood education and child care policy.

[ top ]

1.3 Review of literature

This review draws on a comprehensive search of relevant international literature and aims to identify the major studies conducted in this broad area. It describes evidence on the outcomes of non-parental child care and early childhood education programs on children’s cognitive, early learning, language and socio-emotional development. The review also summarises the key characteristics of child care and early education that may affect such outcomes, and the wider family circumstances of children in child care and early education programs.

The international literature search identified studies relating to three major types of early childhood experiences:

Selection of studies included in the review

This review focuses on non-parental child care and early education programs prior to school. It does not summarise the predominantly United States (US) literature on the effects of high quality, early childhood programs for disadvantaged children, which are typically situated in child care centres or preschools, and include the Perry Pre-School Project (Schweinhart, Barnes & Weikart 1993), the Abecedarian full-day pre-school program (Campbell et al. 2002; Ramey, Campbell & Blair 1998), the Chicago Parent–Child Centres (Reynolds & Temple 1998; Reynolds et al. 2001) and the Head Start and Early Head Start (Administration for Children and Families, US Department of Health and Human Services 2002, 2003), as it is difficult to generalise these programs to broadly available child care and early education programs. Rather, the review focuses on child care and preschool programs of ‘naturally varying’ quality, which conform better to the Australian context.

The subset of studies included in this review involves contemporary cohorts only, rather than cohorts that have now reached adulthood,2 once again because they are more likely to generalise to the current child care context in Australia. The sample of studies was also limited to those large enough (more than 100 cases) and rigorous enough to use statistical analyses that control for parental, family and individual characteristics in assessing effects on children. A brief description of the reviewed studies is provided in Appendix A, accompanied by a summary of the major findings in relation to child outcomes.

Non-parental child care

Scope of existing studies

There are relatively few rigorous studies of child care that involve contemporary cohorts of children and their participation in ‘typical’ child care arrangements, that is, formal care as well as informal care provided by relatives and friends. Three, large-scale (n>1,000) prospective longitudinal studies have been designed specifically to measure child care effects. Two of these are located in the US and one in the United Kingdom (UK). The US studies are the National Institute for Child Health and Development (NICHD) Study of Early Child Care and Youth Development (NICHD–ECC) and the Early Childhood Longitudinal Study, Birth Cohort (ECLS–B). The UK study is the Families, Children and Child Care Study (FCCC). Of these, the ECLS–B is the only study that involves a nationally representative sample.

This review includes other, moderately large longitudinal studies of child care; however, the scope of these varies. In some, the study’s focus is limited to formal child care centres or family day care services. These include the Cost Quality and Child Outcomes in Child Care Centres Study (CQO) (United States), the National Child Care Staffing Study (NCCSS) (United States), Child Care Choices (CCC) (Australia), the Competent Children, Competent Learners (CCCL) Project (New Zealand), the Göteborg Study (Sweden), the Bermuda Study (Bermuda) and the Japanese Study of Extended Child Care (Japan). In others, the study describes outcomes for a range of typical care arrangements, such as the Sydney Family Development Project (Australia), the Stockholm Study (Sweden), the Norwegian Study of Maternal Employment and Day Care (Norway) and the Study of Welfare Reform (California, US). We also include cross-sectional studies relating to relationships between day care and attachment (The Haifa Study of Early Care) and cortisol activity (US Study of Child Care Cortisol Activity; West Australia Child Care Study), as well as the Early Childhood Mental Health Program Study (Washington, US), which employed a retrospective study design.

Child care effects have also been studied using large-scale national cohort studies, including the National Longitudinal Study of Children and Youth (NLSCY, Canada), the Longitudinal Study of Child Development in Quebec (LSCDQ, Quebec) and the Netherlands Twin Register (Netherlands). In addition, the UK Effective Provision of Pre-School Education (EPPE) project (described in the preschool section of this review) provides limited data relevant to child care and child outcomes.

Child care effects and children’s development: effect size

To understand whether use of child care and other types of early childhood programs are related to meaningful differences in children’s development, studies have begun to express findings in terms of effect sizes (NICHD–ECC 2006). Effect sizes estimates are computed from significance tests and are expressed in standard units such that –1 indicates a strong negative association, 0 indicates no association and +1 indicates a strong positive association. Some scholars have dismissed the effect of child care on children’s development as small, when compared to the effects of family background. However, there is much argument concerning the meaning of the effect sizes from different studies, especially in naturalistic studies where many factors influence the conditions and behaviours observed.

Effect size varies by family type (being higher for more disadvantaged families) and by country (being higher where child care is more variable in quality). In the US, for example, follow ups of the 2 year olds, 3 year olds and 4 and a half year olds in the NICHD-ECC study demonstrated modest child care effects (d<0.40). In contrast, parenting effects were described as moderate to large (d=0.50 to 1.2) (NICHD–ECC 2006). Depending on the age of the child and type of outcome, child care effects were between one-third to one-half the size of family environment effects. The most consistent effects were seen for observed care quality and amount of care. In the EPPE study in the UK, spending a longer time in higher quality preschool education generated similar effect sizes to having higher maternal education and higher family income (Sammons et al. 2003). In Australia, Ungerer and colleagues (2006) report moderate effect sizes for amount of child care and aspects of child social development.

Child care effects and children’s cognitive and language development

Using data from the Stockholm Longitudinal Study, Andersson (1989, 1992) reported that children who entered child care in the second half of their first year scored significantly better on standardised measures of cognitive ability and teacher ratings of academic achievement at both 8 and 13 years of age, even after controlling for differences in family backgrounds. The CQO study also found that children who entered child care in the second half of their first year did better on measures of cognitive or language performance, with other factors controlled (ed. Helburn 1995). In contrast, the Bermudian study (McCartney et al. 1985) found that mother’s employment status in the first year of the child’s life made no difference (controlling for family background). Similarly, the analysis of the NLSCY data conducted by Lefebvre and Merrigan (2002) found negligible and insignificant impacts of child care on the motor, social and language development for preschoolers.

Child care effects and children’s socio-emotional development

A major concern of child care research conducted in the US since the 1970s has been whether regular separations between infants and mothers would cause attachment insecurity among infants in child care. The NICHD–ECC provides evidence that early child care (in the first year of a child’s life) does not necessarily harm the formation of the mother–child (or adult–child) attachment bonds that are necessary for healthy infant development (NICHD–ECC 1997, 2001). Similar findings were reported for Australian children (Harrison & Ungerer 1997, 2000). The primary influence on attachment derives from the sensitivity of the mother (adult) in the relationship (Harrison & Ungerer 2002). In the NICHD–ECC study, the amount of child care led to attachment insecurity at 15 and 36 months only if the children were in child care for more than 10 hours a week and if the mother was also highly insensitive towards the child. For mothers providing sensitive care, non-maternal child care did not pose a threat to attachment security of the child.

In relation to the impact of child care on behaviour problems, a recent analysis of the Netherlands Twin registry dataset (Van Beijsterveldt, Hudziak & Boomsma 2005) showed that at the age of 3 years, children with non-parental child care experiences had more externalising problems than children with exclusive parental care. Overall, however, the effect sizes for child care variables were small (effect sizes were between 0.12 and 0.23).

Child care effects and children’s stress hormone (cortisol) levels

Several studies have shown that children in child care record higher levels of the stress hormone cortisol than children who are looked after at home (Dettling, Gunnar & Donzella 1999; Tout et al. 1998; Watamura, Donzella & Alwin 2003). Moreover, while cortisol levels are normally high at the beginning of the day and decrease over the course of the day, the levels recorded when children attended centre-based child care tended not to drop as they do when children were at home. This suggests children attending child care remain in a high state of arousal or stress for longer periods. The pattern of elevated cortisol throughout the day is higher among toddlers, particularly those with more fearful or difficult temperaments (Dettling et al. 2000; Gunnar et al. 1997; Gunnar et al. 2003; Watamura, Donzella & Alwin 2003).

There is evidence to suggest, however, that child care programs in the ‘good to excellent’ range of quality, where teachers are available to respond appropriately to stress, may buffer cortisol release. A study of family-based child care found that children’s stress levels do not rise in settings where they receive a lot of attention, support and guidance from the care provider (Dettling et al. 2000). Similarly, in the West Australian cortisol study (Sims, Guilfoyle & Parry 2006), cortisol levels of children attending child care programs rated as ‘high quality’ on the National Childcare Accreditation Council’s Quality Improvement and Accreditation System ratings decreased across the day, whereas cortisol levels of children attending programs rated as ‘unsatisfactory’ increased.

The work of Gunnar, her colleagues and others has found that negative behavioural effects of child care appear to be related to the pattern of cortisol levels in children over the day. It is suggested that toddlers and preschoolers learning to negotiate with peers may experience child care centres as socially demanding and stressful. Extensive hours in child care and/or the inability of parents and teachers to reduce levels of stress experienced in child care may keep children in a stressed/aroused state.

Child care effects and children’s physical health

Few studies have assessed children’s general physical health in relation to child care, although it is generally accepted that children who attend child care centres are exposed more often to communicable diseases than children in exclusive parental or family care. The NICHD–ECC (ed. 2005) study recorded longitudinally three aspects of child health: ear infections, gastrointestinal infections and upper respiratory tract infections. Findings confirmed that children aged 3 years to 4 years, 6 months in group care (where there are more than six other children) were between 1.4 and two times more likely to be ill than children cared for at home or in small group settings. The rate of infection was not related to the number of hours per week of care. Contrary to expectations, having been in child care before the age of 2 did not reduce the rate of infection at age 3 to 4 years. This study also examined rates of infection from birth in relation to child outcomes at age 3 years and found that the effect of child care on infection was largely unrelated to child behaviour, language ability and school readiness.

Characteristics of child care and child outcomes

A table from the NICHD–ECC study (ed. 2005; p. 32), reproduced as Table 1, summarises the findings for children in the first three years of life on the scope and nature of the relationship between child care and six major aspects of children’s development, after controlling for child and family factors. Effects for quality were consistently identified for three of the six outcomes, despite being tested at differing ages and using different specifications of assessment (indicated by ‘consistent’) and were identified under some conditions for two of the other six outcomes (indicated by ‘sometimes’). In contrast, amount, type and stability of care were less consistently identified as predictors of child outcomes. The nature and direction of these effects are reviewed in the following sections.

Table Description

Table 1: NICHD-ECC study (2005): summary of findings
Child care character Attachment Parent–child relationships Non-compliance in child care Problem behaviours Cognitive development and school readiness Language development
Quality Sometimes Sometimes   Consistent Consistent Consistent
Amount Sometimes Sometimes   Sometimes    
Type     Sometimes Sometimes Consistent Consistent
Stability Sometimes   Sometimes      

Child care quality and cognitive/language outcomes

The studies included in this review demonstrate the critical influence of child care quality on a range of developmental outcomes. Higher quality care has positive developmental effects on cognitive and language development, and on problem behaviours.

The NICHD–ECC and the Bermuda study suggest the verbal environment of the child care setting is of central importance to cognitive and language outcomes (McCartney 1984; NICHD–ECC 2000a). Children enrolled in high quality centres in the Bermuda study scored better on language development and were more considerate and sociable between 3 and 5 years of age. However, the effects of early quality of care no longer appeared at 5, 6, 7 and 8 years, with family background and maternal IQ being better predictors at these ages (Chin-Quee & Scarr 1994).

Positive developmental outcomes were reported in the NICHD–ECC (1999) study when child care centres met professional guidelines for staff–child ratios, group size and care giver characteristics (controlling for maternal education and parenting quality). At 4 years, 6 months, larger group sizes and lower care giver education were separately associated with lower academic achievement and lower cognitive development, controlling for family factors and children’s prior cognitive functioning (NICHD–ECC & Duncan 2003).

Other studies accord with these findings. High quality child care practices were related to improved cognitive outcomes in the CQO study (Peisner-Feinberg et al. 2001). Similar relationships were found between regulatory standards (for example, child–staff ratio or care giver education and specialised training) of home-based child care, defined as ‘where the care provider received payment for child care’ and cared for ‘at least two children including the study child (p. 61), and developmental outcomes (Clarke-Stewart et al. 2002). In the Göteborg Child Care Study, quality of service impacted on the verbal and mathematical abilities of the children who had spent three or more years in child care (Broberg et al. 1997). Likewise, the New Zealand children in the CCCL project—in child care rated as low quality—were less likely than those using child care rated as high quality to be engaging in exploration, or to have extended language in interactions with adults (Wylie, Thompson & Kerslake Hendricks 1996). The Californian Welfare Reform Project Study (Loeb et al. 2004) also showed that child care quality positively affected children’s cognitive and language development. Children displayed stronger cognitive growth when their care givers were more sensitive and responsive.

Child care quality and socio-emotional development

Quality child care indicators have also been consistently related to socio-emotional outcomes among children. In the CQO study, for example, the closeness of the child–teacher relationship was related to adaptive social development (Peisner-Feinberg et al. 2001). The National Child Care Staffing Study also suggested that children who had experienced a history of poor quality care (measured by structural features at 18, 24, 30 and 36 months) were rated as more difficult by preschool teachers and more hostile by kindergarten teachers (Howes 2000; Howes et al. 1988; Howes & Hamilton 1993). Prior to 3 years, 6 months, the quality of child care also shaped children’s social skills in the Göteborg Child Care Study (Campbell, Lamb & Hwang 2000). In the CCCL project, children in low quality child care programs were less likely than others to be engaging in sophisticated play. More aggressive behaviour was observed among these children as well (Wylie, Thompson & Kerslake Hendricks 1996). Children in the Californian Welfare Reform Project Study (Loeb et al. 2004) also displayed stronger social development when their providers had educational levels beyond a high school diploma.

Type of child care and cognitive/language outcomes

Studies using the NICHD–ECC dataset have been better able to separate the effect of child care type from other child care and family effects than studies using any other data set because of the extensive list of controls available. At ages 24, 36 and 54 months (NICHD–ECC 2000a, 2000b, 2002; NICHD–ECC & Duncan 2003), children spending more time in child care centres had higher cognitive and language scores after controlling for family background differences and the quality and amount of child care received. In the NICHD–ECC and Duncan analysis (2003), the measured effect size for cognitive test scores was moderate (0.27). The NICHD–ECC data also suggest that cumulative experience in high quality, centre-based care starting in the second year of life (age 1) may be particularly beneficial for children. This appears to be related to the amount of language stimulation from teachers who have more education and more specialised training in Early Childhood Education (NICHD–ECC 2000a).

In Canada, the NLSCY studies provide a mixed report of the impact of child care centres on cognitive development for preschool children. Lefebvre and Merrigan (2002) found that earlier child care attendance had no significant effect on cognitive development at 4 or 5 years. However, Kohen, Hertzman and Willms (2002) reported that child care centre attendance had a positive impact on language development, and Lipps and Yiptong-Avila (1999) found positive effects on maths scores. These variations in findings may reflect the limitations of the NLSCY data in documenting types and quality of child care experiences.

In Australia, the Sydney Family Development Project (SFDP) (Harrison & Ungerer 2000; Love et al. 2003) showed that competence in learning (such as task orientation, creativity, less learning difficulty) in the first year of school was associated with attendance at formal, regulated care (long day care or family day care) rather than informal, unregulated care during children’s first 2 and a half years.

In the Californian Welfare Reform Project Study (Loeb et al. 2004), children who attended centre-based child care programs performed better on tests of cognitive development than those looked after by friends or family. These developmental effects were moderately strong, especially for measures of school readiness. No consistent differences in cognitive development were found between children who attended licensed, family child care homes and those in other home-based settings. However, children attending family child care homes demonstrated more behavioural problems.

Type of child care and socio-emotional outcomes

Australian research, based on the SFDP, found that infant–mother attachment was more likely to be secure in the group cared for in formal care settings, with the best outcomes being noted for babies in family day care (Harrison & Ungerer 1997). In contrast, when the quality of centre-based programs is poor, as in the Haifa Study of Early Care, infants using centre-based care were significantly more likely to become insecurely attached to their mothers compared with infants who were in different forms of home-based care (Sagi et al. 2002).

Stability and multiplicity of child care and child outcomes

Stability of care (changes in care arrangements or care givers that children experience over time and the use of multiple concurrent child care arrangements) has been investigated only relatively recently. The NICHD–ECC (ed. 2005), NLSCY (Kohen, Hertzman & Willms 2002) and Child Care Choices (Bowes et al. 2004, 2009) have indicated that many infants and toddlers experience high rates of change of child care arrangements, and that multiplicity is a common experience for many young children.

An analysis of the NICHD–ECC data, conducted by Tran and Weinraub (2006), showed that multiple care arrangements involving family members positively predicted language comprehension, whereas multiple care involving a mix of family and non-relative care givers negatively predicted language comprehension. These effects, however, appeared to be influenced by the quality of the primary care arrangement—under conditions of low or moderate quality in the primary care environment, multiple care was associated with lower language scores, whereas under conditions of high quality in the primary care environment, multiple care was associated with higher language comprehension scores.

In the Australian CCC study (Wise et al. 2005), children who attended two or more care arrangements involving regulated long day care centres or family day care plus care provided by a family member tended to have higher scores on global language at preschool-age than children who only attended long day care centres.

For behavioural outcomes, a longitudinal analysis of the SFDP study (Harrison & Ungerer 2000; Love et al. 2003) showed that children who had had more changes in their care arrangements from birth to age 6 were rated by their teachers as less well adjusted (showed more behaviour problems) at school. Similar findings were noted in the CCC study (Bowes et al. 2009): children with a history of having more changes to their care arrangements were rated by teachers as having more behavioural difficulties. This study also noted that children who attending multiple child care arrangements were rated as being less pro-social with peers and having more conflicted relationships with their teachers.

Amount of child care/age at initiation and cognitive/language development

Results are mixed relating to the amount of care, age at initiation and cognitive/language development. The Stockholm Study (Andersson 1989, 1992) reported that children who entered child care in the second half of their first year scored significantly better on measures of cognitive ability and teacher/carer-rated academic achievement at age 8 and 13. In contrast, an analysis of the NLSCY conducted by Gagne (2003) found no correlation between school readiness and the total number of hours spent in child care.

In Australia, Harrison and Ungerer (2000) and Love and colleagues (2003) reported that teachers’ ratings of children’s learning capabilities in the first year of school were lower for children who had received longer hours of care before 3 years of age. Similarly, Bowes and colleagues (2009) have reported an association between longer hours of child care and poorer adjustment to the academic and learning demands of school.

Amount of child care/age at initiation and socio-emotional development

Analyses of several studies have found that extensive amounts of time spent in child care, especially during infancy, are associated with poorer behavioural outcomes. The NICHD–ECC (2001, 2003), for example, found evidence that more hours in child care were associated with behavioural problems, at 2 years, 4 years, 6 months, and in kindergarten. At age 2, more hours in care were associated with more negative interactions with peers and more behaviour problems as reported by care givers, and less social competence as reported by mothers. At 4 years, 6 months, more hours were associated with more negative play, less social competence and more externalising behaviours, and in kindergarten, with more externalising behaviour problems and teacher–child conflict. These effects were moderated by the quality of child care and the quality of parenting, but were still statistically significant and quantitatively important after controlling for these influences.

In the EPPE study, high levels of ‘group care’ before age 3 and especially before age 2 (and entry to preschool) were associated with higher levels of anti-social behaviour at age 3. However, when children showing high levels of anti-social behaviour at age 3 attended quality preschool programs between 3 and 5 years of age, their level of anti-social behaviour decreased (Sylva et al. 2003).

Using data from the NLSCY, Belsky (1986, 1988a, 1988b, 2001) has argued that extensive hours of child care when children are young places them at risk of socio-emotional problems. In contrast, Borge et al. (2004), analysing data from 2 to 3 year olds from the NLSCY, found that aggression was significantly more common in children looked after by their own mothers than those attending group day care. Strong social selection associated with family risk was found, not only in the sample as a whole, but even within the high-risk subsample. However, after accounting for social selection, physical aggression was significantly more common in children from high-risk families who were looked after by their own parents. No such difference was evident in the majority (84 per cent) of children from low-risk families.

Several studies did not find a relationship between quantity of child care and negative behavioural outcomes. Findings relating to the long-term effects of child care quantity from a recent analysis of the Netherlands Twin registry (Van Beijsterveldt, Hudzaik & Boomsma 2005) were mixed and only significant for mothers’ ratings and for children from low socioeconomic families. Children in child care for larger amounts of time did not show more behaviour problems. The Washington study of mental and socio-emotional adjustment (Bornstein et al. 2006) also did not find a relationship between hours of child care and mental development or socio-emotional adjustment. Finally, children in the Stockholm Study who entered child care at an early age were rated more positively on social–personal attributes by their teachers than children who had entered child care at a later stage or who had home care (Andersson 1989, 1992).

In Australia, a recent report by Bowes and colleagues (2009) found that amount of child care in early childhood was a frequent predictor of adjustment outcomes at school-age. Longer hours in formal child care settings (particularly long day care centres) were associated with teachers’ ratings of poorer pro-social behaviour, more socio-emotional difficulties, and more conflicted relationships with teachers.

Effects on children of interactions between child care and family factors

The literature on the effects of early child care remains ambiguous. This underscores the need to focus on the relative impact of parental and non-parental care and how these types of care intersect. Increasingly, studies are considering whether child and family characteristics moderate the influence of child care experiences by examining the interactive effects between child care measures and background characteristics.

On the differential impact of child care on children from disadvantaged backgrounds, the NICHD–ECC study did not find evidence that quality and other characteristics of child care have differential effects across children from different income groups (NICHD–ECC 2000a, 2001, 2003). The authors conclude that all children benefit regardless of family background (NICHD–ECC 2000a). The positive effect of sensitive, stimulating care giving on the development of cognitive and social skills, irrespective of whether children were African-American or English-speaking, Latino or white, was apparent in both the NICHD–ECC and the CQO studies (Burchinal & Cryer 2003).

Moderating influences of family characteristics were observed in the CQO study for some outcomes. Effect sizes, though modest in most cases, were stronger for children from low-income families, indicating stronger positive effects of good quality child care for children from more at-risk backgrounds. The LSCDQ also found that participation in child care programs protected children living in high-risk, disadvantaged families (Borge et al. 2004).

On the other hand, preschoolers from NLSCY families with above average parenting skills and higher levels of education had slightly better cognitive outcomes if they had never been in child care (Gagne 2003). Similarly, in the CQO study there was no evidence that children from more advantaged families were buffered from the effects of poor quality care (Peisner-Feinberg et al. 2001).

Early childhood education services (preschool)

Scope of existing studies

Few studies have examined the effects of preschool education, as opposed to high quality preschool early intervention programs, on children’s outcomes. The reviewed studies include: the US Early Childhood Longitudinal Study–Kindergarten Cohort (ECLS–K), which aims to provide data about the effects of a wide range of family, school, community and individual variables on children’s development, early learning and early school performance; the EPPE, which is the first national longitudinal study of young children’s development (between 3 and 7 years); and the Oklahoma universal, high quality pre-kindergarten program for 4 year olds, which was reported on in a study of 1,843 pre-kindergarten children enrolled in the Tulsa, Oklahoma public schools. Some findings on preschool effects are also available from analyses of the NLSCY and NICHD–ECC datasets.

Effects of preschool programs

Analyses of the EPPE program showed, at the time of entry into primary schooling, that preschool experience enhances both intellectual and social development. The longer children had attended preschool, the greater the intellectual benefits and social abilities (Sylva et al. 2003). Preschool attendance also reduced the rate of risk of special education needs from one in three children to one in five children. Analysis of the ECLS–K dataset conducted by Magnuson and colleagues (Magnuson et al. 2003; Magnuson, Ruhm & Waldfogel 2004) showed that children attending preschool programs had better literacy and numeracy skills at ages 5 and 6, with greater effects for children from disadvantaged backgrounds. Similarly, analyses of data on approximately 1,300 children involved in the NICHD–ECC study (NICHD–ECC & Duncan 2003) suggested that attendance in a preschool program between 3 and 5 years produces higher cognitive scores after age 5 years.

Using the NLSCY data, Pagani and colleagues (2003) compared areas of Canada with and without preschool programs for 4 years olds and found no advantage for children on various behavioural outcomes. A more positive picture emerges from the limited research on studies of the effects of Oklahoma’s universal ‘pre-K’ program for 4 year olds, which controlled for selection effects. Gormley Jr and colleagues (2005) found positive effects on children’s literacy and problem-solving development across all income brackets, irrespective of ethnic background.

The studies reviewed also suggest that children who start school ahead of others in academic achievement tend to stay ahead. Studies using NLSCY data, for example, found that 4 and 5 year-old children entering the first year of school (kindergarten) with low vocabulary scores were more likely to have poor reading scores at ages 8 and 9 years (Hoddinott, Lethbridge & Phipps 2002).

Quality of preschool programs and child outcomes

The available evidence indicates that quality of preschool is related to longer-term outcomes. For example, using data collected in the CQO study, Howes (1988) found relations between dimensions of quality (teacher training, low child–adult ratio, group size of less than 25 children, planned curriculum and adequate physical space) of the preschool program at 3 years of age and a child’s functioning in first grade. After controlling for family factors, children in programs that met more of the quality dimensions had fewer behaviour problems and better work habits.

Type of preschool program and child outcomes

Findings from the EPPE study (Sylva et al. 2003) suggested that outcomes were better when children attended nursery schools and preschool settings that combined care and education. However, the effects of type and quality are difficult to disentangle since these types of preschool education programs also showed higher scores on observed quality.

Amount of preschool and child outcomes

Analyses of the EPPE data suggested that children attending preschool programs benefited cognitively and socially, irrespective of whether they attended a half-day or full-day program (Sylva et al. 2003). The ECLS–K data also suggested that children who start centre-based child care earlier than age 3 make greater cognitive gains than those who start at an older age.

Effects on children of interactions between preschool program factors and family factors

There is some evidence to suggest that children from more disadvantaged backgrounds benefit most from universally accessible early education (preschool programs). Analysing the NLSCY dataset, Kohen, Lipps and Hertzman (2006) reported that the children whose mothers had low levels of education and who participated in pre-school programs before formal schooling rated higher on teacher-reported measures of competence and academic skills than those who did not participate in a pre-school program. Concomitantly, parent-reported behaviour problems were lower among the children who had participated in an early childhood education program. In the ECLS–K study, children from low-income families achieved greater cognitive gains when attending centres for more than 30 hours a week, whereas this did not apply for children from high-income backgrounds (Loeb et al. 2005). Starting earlier was, however, related to more ‘worried/antisocial’ behaviour as rated by teachers, and more than 30 hours in centres per week negatively affected social outcomes among white and African-American children (but not Hispanic children). Analyses of the EPPE dataset showed that children from some ethnic minority groups made greater progress during preschool than white children or those for whom English was a first language (Sylva et al. 2003).

[ top ]

1.4 Discussion

Child care and early childhood education in Australia is a complex phenomenon. Within the broad categories of formal and informal care and early education, there are many types and combinations of care that families use for their young children and these vary by the age of the child. To study the impact of child care/early education on children’s development requires a rigorous research design that accounts for the connections between care and diverse family circumstances, child characteristics and community resources. Assessing possible relations between child care/early education and children’s development requires complex statistical techniques accounting for the effects of family, child and community when analysing possible effects. Results from two decades of international research suggest that the effect sizes for child care are small to modest, when compared to the relatively larger effects of family factors such as parenting and family socioeconomic status.

Research reviewed in this section presents mixed evidence for the links between child care and developmental and health outcomes. This is partly explained by the country in which the study is located, and the regulatory context that this implies. As noted by Love and colleagues (2003), ‘the generalizability of … findings may hinge on the context in which those results were obtained’ (p. 1021). For example, the external regulatory system applied to child care services is an important determinant of the quality of care provided. This is relevant when studying formal child care services in Australia, where responsibility for regulating early childhood education and care (ECEC) services is shared between the Australian Government and state/territory governments and where regulatory arrangements differ for ECEC sectors (Council of Australian Governments’ (COAG) Productivity Agenda Working Group 2008).

Also, the results reported in much of the research involve child outcomes measured beyond infancy. Relatively few studies have examined concurrent relationships between the infant child care experiences and development. Similarly, studies of children’s readiness for school or achievement in school have tended to draw on information about child care before school. Large-scale contemporary research, including LSAC, is primarily longitudinal in nature. However, this report, which deals only with Wave 1 data, is limited in its ability to assess long-term effects.

Four interrelated aspects of children’s child care and early education were reviewed—quality, type, amount and stability. The impact of each of these is summarised here.

Quality

Quality is tied to structural features, such as carer/teacher qualifications, group size, numbers of adults, physical space and resources, and process features such as language stimulation, adult–child interaction and curriculum. Results from a number of studies and locations confirm that quality is important for development, on its own and as a moderator of other factors affecting child outcomes. Better quality programs are generally understood to impact positively on children’s language, cognitive and behavioural development, and social competence with peers and adults. Studies also show that the greater the exposure to better quality programs, the greater the effect, particularly in preparing children for school.

Type

Type of care/early education, because of its association with structural features of the setting, can be closely linked to quality. However, the type of care/early education setting may have different effects at different ages. For example, large group centre-based care, as opposed to small group home-based care, has been linked to more negative outcomes, particularly for very young children. On the other hand, experience in centre-based early education programs for 3 and 4 year olds has been linked to better cognitive outcomes in early years of school.

Amount

Consistent associations have been reported between child behaviour problems and the amount of non-parental child care received. Two complementary aspects of care have been found to be detrimental—an earlier age of entry to care along with longer weekly hours. On the other hand, longer hours or more overall time in high quality care/early education is associated with benefits, such as cognitive gain, for children growing up in circumstances of disadvantage.

Stability

Stability of child care arrangements over time and the number of concurrent child care arrangements are less well represented in the research. In general, more unstable, changeable care is thought to be a risk factor in children’s socio-emotional development. By contrast, ‘multicare’—involving two child care arrangements and often including a small home-based setting with relatives—has been related to positive outcomes for social and cognitive competence.

[ top ]