National Academies Press: OpenBook

Who Cares for America's Children? (1990)

Chapter: II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care

« Previous: 2 Trends in Work, Family, and Child Care
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 43
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 44
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 45
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 46
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 47
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 48
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 49
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 50
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 51
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 52
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 53
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 54
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 55
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 56
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 57
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 58
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 59
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 60
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 61
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 62
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 63
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 64
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 65
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 66
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 67
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 68
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 69
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 70
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 71
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 72
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 73
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 74
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 75
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 76
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 77
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 78
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 79
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 80
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 81
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 82
Suggested Citation:"II CHILD CARE AND CHILD DEVELOPMENT3 The Effects of Child Care." National Research Council. 1990. Who Cares for America's Children?. Washington, DC: The National Academies Press. doi: 10.17226/1339.
×
Page 83

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

II CHld Carc and Chad Dcvelopmcnt

3 The Effects of Child Care Chapters 3 through 5 review what has been learned about the lives of children in child care from research, clinical practice, and work with young children. In Chapter 3 we trace successive phases of child care research, which have moved to increasingly complex and fruitful questions about child care. In Chapter 4 we highlight the particular dimensions of child care quality (e.g., group size, ratio) that are most important to children's development. And in Chapter 5 we review what is known about how child care can support children's physical health and psychological development. PROCESSES OF CHILD DEVELOPMENT Before turning to these issues of child care and children's develop- ment, however, it is useful to outline briefly several basic principles about development that underlie the ensuing discussion. These principles emerge from and reflect important areas of agreement in different scientific dis- ciplines that shed light on children's development, notably developmental psychology, clinical work with children, and research in early education. First, children's development is multiply determined: by sources within the child, such as temperament, neurological integrity, and impairment; by factors in the child's immediate environment, such as quality of relationships and interactions with parents and quality of out-of-home care; and by factors in the child's larger social environment, including the immediate neighborhood and the broader culture. These factors do not operate separately but interact in a complex fashion to influence developmental outcomes. Child care must therefore be viewed as one of many sources of influence on children's development and one that interacts in complex ways with numerous others. 45

46 WHO CARES FOR AMERICA)S CHILDREN? Second, children are not only influenced by their immediate and broader environments, but also shape these environments. Development reflects the transactional processes-or mutual influences of child and environment. In child care settings, for example, children are influenced by caregiver and peer behaviors, and they also selectively relate to certain teachers and peers. Third, children's development involves the biological, cognitive, and socioemotional domains. Although development in these domains is in- terrelated, progress across domains does not occur uniformly. Rather, children may advance or lag in one domain but not others. It is therefore important to consider the development of children in child care in specific domains, rather than to view patterns of influence as equally affecting all domains. Fourth, children's physiological, cognitive, and socioemotional needs differ markedly by developmental level. Consideration of the age of a child is fundamental to understanding the differing needs of children in child care settings.) Finally, the nature of environmental influences on children is best thought of as probabilistic. Recent research and practice with children posits risk factors, conditions or events that increase the probability of negative or less than optimal developmental outcomes, and protective factors, conditions or events that increase the probability of positive or optimal developmental outcomes. Risk and protective factors are thought to influence development most often by interacting with other sources of influence on a child's development. Child care could function as a protective factor, a risk factor, or a relatively neutral factor for particular developmental domains. In short, the relationship between child care and child development is complex. In light of this complexity, it is not surprising that there are no perfect studies and few that are conclusive about the relationship between child care and child development. Nonetheless, the cumulative weight of evidence from empirical studies, clinical work, and professional practice is sufficient to draw some conclusions that can serve as provisional guides to program and policy. The rest of this chapter and the next two review that evidence. ~ In this chapter, and indeed throughout this volume, infants refers to children in the first year of life, toddlers to those 13 to 36 months, preschool age to those 3 to 5 years old, and school age to those 6 years or older. We note, however, that there is some variation across researchers and legislators in the precise demarcations used for these developmental periods. Where a particular study, author, standard, or regulation uses age demarcations that differ from ours, we note the discrepancy and specify the intended age of the children.

THE EFFECTS OF CHILD CARE EVOLUTION OF CHILD CARE RESEARCH 47 Marked social changes that impinge on the lives of families and chil- dren have often sparked an initial wave of psychological research that asks, "Is this harmful to children?" It has been the case repeatedly that this initial wave of research, which might be called the "alarm phase," does not address questions of sufficient subtlety or complexity to illuminate the impact of the social change. Typically, in the evolution of the research, the alarm phase gives way to a second research phase that examines the demographic shift in a more differentiated manner. The research on maternal employment (a "sister" literature to that on child care that is not restricted to young children receiving a particu- lar form of supplemental care) illustrates this evolution. Researchers and practitioners responded decades ago to the striking increases in rates of maternal employment by asking whether or not children were adversely affected by daily separations and nonmaternal care. But the great complex- ity in the research findings from those early studies indicated that it was not enough to ask about potential harm. Some children clearly benefited from their mothers' employment. Furthermore, it did not appear that ma- ternal employment was a unitary phenomenon with uniform implications (Bronfenbrenner and Crouter, 1982; Hoffman, 1979~. Rather, its impact on children was related to a number of child characteristics (e.g., age and sex), family characteristics (e.g., father involvement with children, mothers' role satisfaction, and extended family support), and factors beyond the family (e.g., culture) (Zaslow et al., in press). Research restricted to a status strategy mother is or is not employed- gave way to research focusing on mechanisms and processes, the factors associated with differing outcomes of maternal employment for children. The research on child care is following a similar pattern of evolution (Belsky, 1984; Phillips, 1988~. A first wave of studies asked whether the increasing rates of participation of young children in family day care and center care was cause for alarm. Did the development of children in supplemental child care differ from that of children cared for by parents? The accumulating evidence, however, has forced a shift in the research focus. Child care was not found to be inherently or inevitably harmful. Rather, it was found to vary greatly as an environment for children, and children's development was linked to the variation. A second wave of child care research, still in progress, examines children's development in light of variations in the environment of child care. According to Phillips (1988), the major focus of a third wave of child care research is beginning to emerge. This phase will view the home and child care environments as linked and mutually influential. Children's development is beginning to be understood in light of experiences across the two care settings.

48 WHO CARES FOR AMERICA'S CHILDREN? Each successive wave of child care research reflects progressively greater conceptual complexity and methodological refinement. Yet the question central to each wave remains important, even as the new ques- tions emerge. That is, the waves of research overlap and complement one another, rather than supplanting each other. It is important to continue to ask how the development of children reared at home and in child care settings differs (the question focal to the first wave of research), even as questions are asked about the implications of poor-quality care in comparison to care that is adequate or of high quality (second wave) and about the joint impact of the child's home and the child care settings (third wave). Although there are methodological weaknesses as well as strengths in each research wave, we view each phase as using research approaches appropriate to the central question being addressed. We begin, however, not with the first wave, but one step further back, in the research on maternal deprivation. ROOTS OF CHILD CARE RESEARCH: MATERNAL DEPRIVATION The alarm phase of research in child care has its roots in the substantial body of clinical and empirical studies of children experiencing maternal de- privation and institutionalization (long or short stays in residential facilities for children). It was not only initially considered possible that child care might belong on a continuum with the institutionalization of children, but actually explicitly stated that this was the case. The influential 1951 World Health Organization Expert Committee on Mental Health concluded on the basis of the work of John Bowlby (1951), that "day nurseries" consti- tuted a form of maternal deprivation with permanent negative effects on children (Rutter, 1981a). Studies of short-term (hospitalization, short-stay residential nursery) and long-term ~ong-stay residential nursery, institutionalization) parent- child separations, many of them rooted in the psychoanalytic tradition, do indicate problems in children's development (e.g., Freud and Burlingham, 1944, 1973; Goldfarb, 1943; Provence and Lipton, 1962; Ribble, 1965; Robertson and Robertson, 1971; Spitz, 1945; Wolkind, 1974~. These prob- lems range from acute distress syndrome, associated with short periods in care, to conduct disorders, problems in forming relationships, and in- tellectual deficits, associated with longer term care (Rutter, 1981a). In particular, `'an institutional upbringing which involves multiple changing caretakers has been shown to lead to important social deficits and prob- lems in interpersonal relationships" (Rutter, 1981a:154~. But does the ac- cumulated evidence regarding weeks or years completely away from parents

THE EFFECTS OF CHILD CARE 49 indicate that child care should be considered together with this grouping of studies, as a form of maternal deprivation? Rutter's (1979, 1981a) reassessments of the theory and research into maternal deprivation are extremely helpful in extracting those elements of early views of maternal deprivation (particularly those of Bowlby) that the research evidence has sustained and those elements that have required revision. According to Rutter, there have been two important and lasting contributions of the early work on maternal deprivation. First, that work identified deplorable conditions in short- and long-term residential settings for children. In response, there have been widespread changes in those conditions. Second, studies responding to that work have resulted in an understanding of experiences that are necessary for young children's normal development. Above all, that research showed that young children need to develop enduring relationships with a limited number of specific individuals, re- lationships that are characterized by affection, reciprocal interaction, and responsiveness to the particular and highly individualized cues of the infant and young child, and that the child's environment has to provide sufficient opportunity for stimulation (Rutter, 1981a, summarizes the evidence). It is in the context of familiar relationships that children make their major developmental advances in communication and understanding in the first years. The growth of language and social understanding depends on the child's social exchanges with familiar, responsive others (Bruner, 1983; Dunn, 1988; Lock 1978; I3revarthen, 1977~. In other respects, other early conclusions on maternal deprivation have required revision in light of subsequent research (Rutter, 1979, 1981a). Per- haps most central, the initial work emphasized the child's need to become attached to the mother in particular, and to be cared for in his or her own home, for healthy subsequent development. For example, Bowlby's early statements viewed the mother-child bond as different from all other rela- tionships (Bowlby, 1951~. Subsequent work has not confirmed this exclusive emphasis on the mother-child relationship. Rather, the evidence indicates a normal tendency for children to form multiple simultaneous attachments (Chibucos and Kail, 1981; Dunn, 1983~. These attachments tend to be hierarchical, with one attachment (often to the mother) most significant to the child (Rutter, 1981b). However, mothers are not the only caregivers who can provide the essential experiences for healthy early development. Children can benefit from '`multiple mothering" if it provides affection, warmth, responsiveness, and stimulation in the context of enduring rela- tionships with a reasonably small number of caregivers (according to Rutter [1981a], four or five caregivers), who have come to know and be able to respond to the child's individual needs and style.

so WHO CARES FOR AMERICA'S CHILDREN? Accordingly, researchers no longer believe that the effects of institu- tionalization on children are solely the result of disruption of the mother- child relationship. In the more complex and differentiated view of insti- tutionalization that has evolved, it is now understood that some of the more serious developmental problems follow not from interruption of a mother-child relationship, but from an initial lack of any such relationship. Furthermore, the effects of residential care vary according to the setting's provision of enduring relationships and stimulation; according to the child's physical status (e.g., nutrition, pre- and postnatal complications); and to the family circumstances surrounding the institutionalization, particularly family discord. Thus, the nature and circumstances of the separation in combination with the characteristics of the child determine its implications rather than the single fact of mother-child separation (Rutter, 1979, 1981a; Wolkind, 1974). Of particular importance in the present context, Rutter's reassessment of the maternal deprivation research concludes that child care does not fall on a continuum with institutionalization (Rutter, 1981a:154~: [where is a world of difference between institutional care without any parental involvement and day care in which the mother remains a key figure who continues to actively participate in looking after the child. For these reasons, little weight can be attached to the results of residential group care as a basis for assessing the probable sequelae of group day care. In sum, infants and young children tend to form several attachments to a small number of selected individuals. A healthy environment for child development does include at least one secure attachment, but does not necessarily require care exclusively by a mother. Rather, environments that provide stable, warm, responsive, and stimulating relationships with several caregivers, if these relationships are limited in number, can encour- age healthy psychological development. Child care differs fundamentally from institutionalization and should not be considered part of the same phenomenon simply on the grounds of mother-child separation. Yet when child care environments fail to provide those elements found to be essential to normal development-and instead involve a large number of caregivers, frequent changes in caregivers, lack of responsiveness to the child as an in- dividual, and lack of social and cognitive stimulation they can be expected to have negative implications for development.

THE EFFECTS OF CHILD CARE FIRST WAVE OF CHILD CARE RESEARCH: DEVELOPMENT OF CHILDREN IN CHILD CARE AND THOSE REARED AT HOME 51 The impetus for the first wave of empirical, systematic child care re- search, then, was the concern that young children would be harmed by daily separations from their mothers. As it became clear that mother-child separations and daily participation in child care did not have the drastic negative implications that the maternal deprivation construct predicted, important questions nevertheless remained: Did the development of chil- dren in supplemental child care differ from that of home-reared children in meaningful if less marked ways? Were there subgroups of children whose development was enhanced by the child care experience? Were there sub- groups of children whose participation in child care was associated with any degree of risk for development? Several comprehensive reviews of the first wave of child care research by Belsky and colleagues (Belsky, 1984; Belsky and Steinberg, 1978), Clarke- Stewart and Fein (1983), and Rutter (1981b) have addressed these questions and provide extensive documentation, summarized below. Consistent with the assumptions about development presented in the preface to these chapters, we assume that the impact of child care differs by domain of development. As such, we present the major findings separately for the areas of intellectual and social development. However, we cannot present findings separately by developmental level because researchers have made surprisingly little effort to differentiate the outcomes for children exposed to child care according to child age. The evidence rests largely on a mosaic of cross-sectional studies, rather than on longitudinal studies aimed at documenting the changes and consistencies in children's needs in child care as they get older. The major exception to this pattern is that the infancy period has been singled out for particular focus. But the years from toddlerhood through school age are rarely discussed in the child care literature with an assumption of changing developmental needs. A further limitation in this wave of child care research in the United States that is important to note is its very heavy reliance on studies of center care, rather than the demographically more prevalent family day care. In this respect much can be learned from the European research, which has more consistently encompassed family day care as well as center care in its attempts to examine the impact of child care experience versus parental care (e.g., Cochran, 1977; Lamb, Hwang, Bookstein, et al., 1988; Lamb, Hwang, Broberg, and Bookstein, 1988~. Intellectual Development On measures of intellectual development, reviews of the evidence conclude that "children in day care centers do as well as those at home

52 WHO CARES FOR AMERICA'S CHILDREN? . . . or that they do better, at least for a time or on some measures" (Clarke-Stewart and Fein, 1983:965~. Social class is an important factor in understanding the findings for cognitive development among children in child care (Belsky, 1984; Belsky and Steinberg, 1978~. Studies of more economically advantaged children in community-based child care find either no differences in cognitive devel- opment related to child care participation or more advanced development among these youngsters. Studies of economically disadvantaged children in high-quality child care intervention programs, however, consistently find more advanced cognitive development in day care children than in home- reared children. These children do not show the declines found for their home-reared counterparts from disadvantaged families on tests of intellec- tual development. Although Belsky's reviews (Belsky, 1984; Belsky and Steinberg, 1978) conclude that, overall, middle-class child care and home-reared children do not differ on indices of intellectual development, Clarke-Stewart and Fein (1983) diverge in their assessment of the research, pointing to indications of superior scores on cognitive indices in some studies of children attending community-based child care programs (e.g., Doyle, 1975; Rubenstein et al., 1981~. Thus, for more economically advantaged children in community- based programs, the most consistent conclusion appears to be that child care attendance does not have negative implications for cognitive development. In children from disadvantaged families, measures of IQ (that are lan- guage dependent) typically decline beginning in the second year of life (see Slaughter [1983] for discussion of this pattern and the types of measures on which it occurs). Reviews of the evidence on early intervention pro- grams for children at risk for this decline indicate that such programs are effective in preventing or slowing the decline. Bryant and Ramey (1987), for example, in an excellent recent review of this evidence, restricted their examination to methodologically adequate experimental studies of early intervention programs and considered the role of child's age at entry, dura- tion and intensity of intervention program, nature of educational activities, and whether the child or parents were primary targets of the intervention. They conclude that program effectiveness was most closely linked with the child's (or family's or both) extent of contact with the intervention pro- gram and that "the most improvement in intellectual development occurs when children attend day care and families receive parent training or other services" (Bryant and Ramey, 1987:71~. Interventions were found to be effective irrespective of whether they were initiated during infancy, early childhood, or the preschool years. And the type of educational emphasis and the curriculum were not closely linked with outcomes. Findings of the early intervention programs indicate that gains on

THE EFFECTS OF CHILD CARE 53 measures of intellectual development are temporary rather than permanent (evidence summarized by Clarke-Stewart and Fein, 1983; Haskins, 1989~. However, longitudinal evaluations of early intervention that include out- come measures in addition to IQ show some persistent effects. For example, in one longitudinal analysis of 10- to 17-year-olds who had attended early intervention programs, although group differences on IQ did not persist, differences on measures of school-related behavior did: children who had participated in intervention programs were less likely to repeat a grade in school and less likely to be referred for special education than those who had not participated (Darlington et al., 1980; Lazar et al., 1982~. Recent research also indicates that when an intervention program is continued into the elementary school years, complementing the child's regular school participation, differences can be sustained. Horacek and colleagues (1987:762) found that "children who participated in both [a] preschool program and [a] school-age support program performed better in school than the group that had only preschool" intervention, even though the intervention at school age was of limited intensity. Thus, children at risk for school failure may benefit most from a combination of early and sustained intervention. Haskins (1989), in reviewing the findings on the impact of early inter- vention programs, urges a distinction between model intervention programs (e.g., those considered in the Bryant and Ramey [1987] review) and Head Start. Haskins points out that both types of programs yield "significant and meaningful gains" on measures of intellectual performance by the end of the first year of intervention, but it is only for the model programs that there is evidence of strong positive effects on the later school-related behavior variables. He suggests several possible explanations for this differ- ence in the findings for model early intervention programs and Head Start. Since data collection has been far more extensive and systematic for the model programs, the long-term effects of Head Start may be undetected by the fewer and less rigorous Head Start outcome studies. However, there are fundamental program differences that could underlie the difference in longer term implications of the two kinds of programs: Head Start en- compasses sites ranging substantially in quality; while model programs are consistently high quality. Furthermore, Head Start selects for participation the most disadvantaged children and families, but draws control groups from those remaining on the waiting list and who are thus relatively less disadvantaged. Work by Lee and colleagues (1988) suggests that particu- larly because of Head Start selection practices the impact of Head Start may be systematically underestimated in studies. Haskins (1989), while apologizing for the tendency of social scientists to call for more research, notes the particular need for methodologically

54 WHO CARES FOR AMERICA'S CHILDREN? rigorous longitudinal studies of Head Start. We concur that Head Start, a federally supported, comprehensive early child development intervention, with strong evidence of short-term benefits in the intellectual domain, should be the focus of carefully planned longitudinal studies that track a broader array of social, emotional, and cognitive outcomes. Social Development Attachment In studying the socioemotional development of children in child care, researchers have been concerned about the nature of mother-child relations. Do children in supplemental care show patterns of attachment similar to those of home-reared children? What happens to children who enter child care during the developmental period (the first year of life), when the attachment to mother is forming? Before turning to the evidence on child care and security of attachment, we briefly consider the attachment construct. The Attachment Construct: Defi~ii'on and Assessment The term "attach- ment," as used in the- psychological research, has its roots in the work of Bowlby (1969, 1973, 1980~. Bowlby stressed two central functions of an infant's enduring relationship with its mother: the provision of a "secure base" from which the infant could explore the environment and the pro- vision of a "haven of safety" to return to when stressed or distressed (see Campos et al. [1983] for a discussion of the evolution of the attachment construct). The security of the infant with the mother, that is the use of the mother both as a secure base and a haven of safety, has been widely evaluated using an assessment known as the Ainsworth "strange situation" (Ainsworth et al., 1978~. In this laboratory situation, infant behavior to- ward its mother is observed in a sequence of eight episodes involving introduction to a novel situation in the presence of the mother, behavior toward a female stranger in the presence and absence of the mother, infant behavior when left alone in the novel setting, and reunion behaviors with the mother. The components of the strange situation (unfamiliar setting, unfamiliar though friendly adult, separation, and reunion) were devised in keeping with Bowlby's view that attachment behaviors would be most readily observed in a context in which the baby is stressed or aroused. The Ainsworth strange-situation assessment, as it has been widely used, distinguishes three qualitatively different patterns of attachment (although recent work raises the possibility of a fourth category). In this assessment, infants rated as securely attached "tend to seek proximity to, and contact

THE EFFECTS OF CHILD Cal 55 with, attachment figures.... Moreover, such infants manifest clear pref- erences for their caretakers over the stranger" (Campos et al., 1983:863~. Approximately two-thirds of middle-class American infants observed in the strange situation (Campos et al., 1983) are categorized as securely attached; while one-third are rated insecurely attached in one of two ways. "~xious- avoidant" infants "conspicuously avoid their caretaker during the reunion episodes, fail to cling when held, and tend to treat the stranger the same way as, or sometimes more positively than, their caretaker" (Campos et al., 1983: 862~. "Resistant" or "ambivalently attached" infants "tend to resist interaction and contact with their caretaker, yet they also manifest contact- and proximity-seeking behavior. Theses infants seek proximity and contact before separation, moreover, which may inhibit their exploration of the novel environment" (Campos et al., 1983:863~. A body of research relates these three patterns of attachment both to antecedents (particularly the nature of the mother-infant interaction) and asks whether the pattern of attachment is predictive of aspects of development. Features of mother-infant interaction related to the differing patterns of attachment are the mother's interest in and availability for interaction with the infant, as well as the emotional tone of interactions. Work by Ainsworth and colleagues with a middle-class sample (summarized in Ainsworth et al., 1978) indicates that mothers of securely attached infants were more effective in soothing and interpreting infant signals, that they participated in more face-to-face interaction with their infants, and that they were more affectionate and emotionally positive with the infants. By contrast, mothers of anxious-avoidant babies were more irritable and rejecting in their interactions, and mothers of resistant infants were more inept and insensitive to signals. Subsequent work has continued to differentiate mother-infant interactions according to pattern of attachment, but the particular differentiating features have not always been identical nor has the differentiation been unambiguous or strong (see evaluation by Campos et al., 1983~. Campos and colleagues (1983) conclude that the evidence that the strange-situation classification predicts developmental status is more robust than that it has unambiguous roots in early interaction. Thus, for example, children categorized as securely attached as infants have been reported to be subsequently more compliant and cooperative (e.g., Matas et al., 1978), more sociable with peers (Pastor, 1981) as well as more competent with peers (Waters et al., 1979), and to engage more effectively in carrying out cognitive tasks (e.g., Matas et al., 1978~. Before asking whether participation in child care is related to security of attachment as assessed in the strange situation, it is important to call attention to concerns expressed in the literature about the use of this as- sessment. These concerns relate both to the strange-situation classifications

56 WHO CARES FOR AMERICA'S CHILDREN? in general and to the application of this procedure specifically to children who have experienced frequent separations from their mothers because of maternal employment (Campos et al., 1983:869~: So widely accepted is the system of Ainsworth et al. (1978) for classifying individual differences in the Strange Situation that few have stopped to ask whether [its] trichotomy constitutes a valid way of clustering individual differences.... Indeed, an evaluation of the research carried out by Campos and colleagues asks (1983:872) whether it is "justifiable to equate 'security of attachment with Strange-Situation classification."' For example, they note a wide range in reports examining the stability of an attachment classification over time, including reports that only approximately one-half of a sample retains the same classification over time (e.g., Thompson et al., 1982~. In addition, they note the possibility that behavior in the (stressful) strange situation may be a reflection not only of the history of mother-infant interactions, but also of the baby's temperamental characteristics, including how easily the infant becomes distressed and is comforted, characteristics that "may be evident long before the attachment relationship is built" (Campos et al., 1983:868~. Campos and colleagues question the original attachment groupings and the manner in which they were determined. Finally, they point to widely divergent proportions of infants categorized as insecure and secure in attachment in studies carried out beyond the United States (e.g., the work of Grossman et al. [1981] reporting a higher proportion of German infants showing anxious-avoidant attachment). Such differences raise the possibility of important cultural or experiential differences among infants in the strange situation, particularly the extent to which the baby finds the experience stressful. In keeping with this perspective on the cross-cultural data, other re- searchers have questioned the equivalence of the strange-situation assess- ment for infants whose mothers have rarely parted from them in comparison with infants who are accustomed to daily departures. Clarke-Stewart (1989) in particular questions whether this procedure is "psychologically equiva- lent" for infants of homemaker and employed mothers. She notes that it is much less likely for the child of an employed mother to see as un- usual or stressful the experiences of a novel play setting, being left by the mother with a female stranger or being comforted by her, and reunion with the mother. Clarke-Stewart (1989:267) concludes that "we need to assess infants' attachment using procedures that are not biased by differential familiarity and potentially differential stressfulness." Because of these questions, the panel regards the data on security of attachment among children who have participated in child care with some caution. The almost exclusive reliance on the strange-situation procedure

THE EFFECTS OF CHILD CARE 57 offers one advantage in that one can look at comparable data across studies (Clarke-Stewart, 1989~. Yet, at the same time, there has been a tendency among some researchers to uncritically accept the validity of the strange- situation classification, even though questions remain about its equivalence for samples with different care histories. Security of Attachment and Participaiion in Day Care The findings on security of attachment and child care participation are best summarized separately for children who begin child care in the first year of life or later. For the somewhat older children, attendance in a child care program does not appear to alter the hierarchy of attachments (Rutter, 1981a,b). Most children do develop attachments to stable caregivers and seem to gain security from their presence (Howes et al., 1988~. Yet, most children consistently prefer proximity, interaction with, and comfort from their mothers (see Rutter [1981b] for evaluation of this evidence). Therefore, daily hours apart do not alter mothers' primary role in the lives of children in child care. For these somewhat older children, the months following entry into child care may involve "transient distress" that manifests itself in the mother-child relationship (Belsky, 1984~. However, beyond this adaptation period, there are no marked differences in the quality of attachment to mother for children in nonparental care and home-reared children in this age range. In evaluating these findings, it is important to recall that having a mother who is a homemaker does not ensure secure attachment or optimal mother-child interactions. As noted above, a nontrivial proportion of toddlers with homemaker mothers show `'insecure" attachments as assessed in the strange situation (Richters and Zahn-Waxler, 1988; Thompson, 1988~. Furthermore, being a homemaker has been found to be associated with depression among working-class women in England (Brown and Harris, 1978), and depression in the mother, in turn, is associated with problems in preschool children. Among children who begin their child care attendance for more than 20 hours per week during the first year of life, researchers now agree that, while a majority show secure attachments to their mothers when tested in the strange situation, a higher proportion of the remainder show anxious-avoidant attachment to their mothers than do home-reared infants (Barglow et al., 1987; Belsky, 1988; Belsky and Rovine, 1988; Schwarz, 1983~. There is agreement that infants who start full-time child care in their first year are more likely to show this pattern of attachment, but there is no consensus about what it means (Belsky, 1988; Clarke-Stewart, 1989; Richters and Zahn-Waxler, 1988; Thompson, 1988~. Some researchers con- tend that it resects an undesirable pattern in the infant-mother relationship

58 WHO CARES FOR AMERICA'S CHILDREN? that may have negative implications for later development (Belsky, 1988; Belsky and Rovine, 1988~; others conclude that the pattern is a manifes- tation of other factors, such as patterns of self-selection, which distinguish families that do and do not use child care for infants, a different but healthy adaptation in infants of employed mothers, or a reflection of the methodological issues noted above. Rutter (1981b) and Clarke-Stewart (1989) both emphasize the possi- bility of self-selection factors distinguishing between families that rely on parental care and those that rely on child care. Perhaps the most revealing studies on psychological factors that differentiate between mothers who choose to be homemakers or choose to be employed have been carried out by Hock and colleagues. For example, Hock and colleagues (1980), looking at a group of mothers of newborns all of whom planned to be homemakers, found differences between those who carried out this plan and those who changed plans and resumed employment. Three months after the births of their children the mothers who resumed employment expressed less positive attitudes about the maternal role and greater diffl- culty with infant fussiness. Hock and colleagues (1984) found that in the newborn period, mothers expecting to be homemakers differed from those expecting to return to employment in terms of how strongly they believed that it is important for babies to be cared for exclusively by their mothers and in their home orientations. Such differences may have implications for the development of infant-mother attachment. For example, the tendency of mothers returning to employment to experience infant fussiness as more aversive may imply that they are less responsive to their infants' distress or less effective responding to it. Both patterns, as noted above, have been linked with the emergence of insecure attachments. Similarly, Crockenberg (1981:862) found that "the adequacy of the mother's social support is clearly and consistently associated with security of infant-mother attachment" and with higher rates of avoidance and anxious attachment when mothers have little social support. Perhaps mothers who remain home with their infants have better networks of support. Indeed, such a factor could contribute to their decisions to be homemakers. Researchers have also raised the possibility that findings of higher rates of anxious-avoidant attachment in infants of employed mothers may simply indicate a healthy adaptation by infants to child care: such infants may show greater autonomy from their mothers or less distress in the assessment situation because they are accustomed to separations (Clarke-Stewart, 1989; Clarke-Stewart and Fein, 1983~. A further possibility (although one that Clarke-Stewart [1989] evaluates as not supported by available data) is that higher rates of insecure attachment in infants who have attended child care are linked to poor-quality early care. Perhaps most important, studies to date have not yet followed day care children showing anxious-avoidant

THE EFFECTS OF CHILD Cal ~9 attachment to assess directly the stability of the pattern or its developmental implications. Stress and Parenting During the First Year The fact that negative findings concerning attachment pertain specifically to children who participate in full-time child care during the first year of a child's life have led some researchers to ask whether stresses unique to this period are indeed taking a toll on parent-infant relations. Research in the United States has long debated the extent of disruption to parents and to the marital relationship when a baby is born (e.g., Dyer, 1963; Hobbs and Cole, 1976; LeMasters, 1957), but it is clear that the birth of a child involves rapid readjustment within the family that is both stressful and positive. Employed women in particular tend to experience an intensification of "role overload" (too much expected given dual roles) and "role conflict" (internal conDict about the relative importance of work and family roles) with the birth of a child (Moen, 1989~. Brazelton (1986) has questioned whether mothers' early return to employment, when added to the stresses of parenting an infant, may limit the time and energy mothers and infants have to establish a pattern of mutual communication and sensitivity to cues. Furthermore, he points to possible obstacles affecting the mother's ability to develop strong positive feelings for and about the infant when she needs to cope with frequent separations and shares her baby's care with another caregiver. v_r~ Observations of parent-infant interaction in the first year of life suggest that infants in middle-class families in which the mother is employed are engaged in somewhat less playful interaction with their parents than infants with homemaker mothers (Zaslow et al., 1989~. Further research suggests that secure attachment may not be used in the same way by infants in families with employed mothers as it is in families with homemaker mothers. Vaughn and colleagues (1985) found that security of attachment in infancy was an excellent predictor of later socioemotional development in children whose mothers were homemakers, but it did not predict later socioemotional competence in children whose mothers returned to work when their babies were very young. Clarke-Stewart (1989) notes several ways in which stresses unique to combining employment and care of an infant might be associated with increased rates of anxious-avoidant attachment. Anxious-avoidant attach- ment, as noted above, may be rooted in a rejecting quality to mother-infant interactions. Clarke-Stewart (1989:270) points out that "the increased stress of handling two full-time lobe work and motherhood, [could] lead to more --he ~ J___7 rejection of every additional burden, including the baby." Alternatively, a perception of rejection by the baby might not be a reflection of the mother's feelings about the baby, but simply her inaccessibility because the

60 WHO CARES FOR AMERICA'S CHILDREN? tasks facing her when she returns home compete for her time and attention with the baby. Clarke-Stewart notes several studies that relate attitudinal or per- sonality factors among employed mothers (e.g., desire for motherhood, psychological integration, anxiety, and dissatisfaction) and security of at- tachment in their infants (Been, 1986; Farber and Egeland, 1982; Owen and Cox, 1988~. It is possible that higher rates of anxious-avoidant attachment among infants in child care are related to higher proportions of employed mothers feeling anxious, stressed, or overburdened. Findings from Sweden point out that providing greater flexibility in employment roles may reduce the stress that many employed mothers experience. Moen (1989) indicates that reports of daily fatigue and psychological stress by mothers in Sweden are related to the recent birth of a first child. The use of parental leave or the reduction of hours of employment to a part-time schedule significantly reduce both indicators of stress. Moen also notes that mothers experience significantly greater stress than fathers in Sweden, despite the availability of employment and leave options. These findings, taken together, have implications for evaluating paren- tal leave policies in the United States. We must distinguish between what the data permit us to say with some certainty, and where there are problems with the evidence. We can say with some certainty that U.S. mothers of infants, who are also employed full time, experience overload and stress. Furthermore, we know that factors reflecting psychological distress among employed mothers are related to the emergence of insecure attachment in their infants. Researchers agree that infants of mothers who resume full-time employment in the first year of their infants' lives show higher rates of anxious-avoidant attachment to their mothers. Finally, there is evidence from Sweden that parental leave or reduction in hours of parental employment can reduce stress in mothers. In the United States, evidence suggests that it is only full-time employment, not part-time work, that is associated with the pattern of anxious-avoidant attachment in infants. Yet there is no research to date that puts these pieces of evidence together to establish paths of influence. That is, there is no research in the United States examining directly the effect of a period of parental leave or reduced hours of employment on stress among mothers, on the quality of mother-infant relations, or on rates of secure and insecure attachment among their infants. Furthermore, as we have noted, there are important questions about the single assessment of the mother-infant relationship that has been widely used and particularly about its use with children who have participated in out-of-home child care. Support for a parental leave policy in the United States at the present time cannot be built on definitive data about its implications for infants, mothers, and families. It can, however, rest on a set of individual findings

THE EFFECTS OF CHILD CARE 61 sufficient to raise concerns about stress in employed-mother families with infants and its possible implications for children. Summary of Finding, on Attachment and Child Care For children beginning child care after the first year of life, there is little indication of differences in the mother-child relationship beyond an initial adaptation period. Children beginning full-time child care within the first year, however, show higher rates of anxious-avoidant attachment to their mothers than other children. The interpretation and implications of this pattern need to be further scrutinized. In particular, it should be a high priority in future research both to examine the use of the strange-situation assessment in infants of employed mothers and to substantially extend assessment of the mother- infant relationship beyond this single measure. Research is also needed on the context in which anxious-avoidant attachment arises among infants of employed mothers; to trace the development of infants of employed mothers with differing attachment ratings; and to directly assess the impact of parental leave on stress among mothers in the United States and on the quality of the mother-infant relationship. Relationships With Peers and Adults Child care researchers have observed and documented the social re- lations of children in child care (as opposed to home-reared children) in two areas other than the mother-child relationship: relations with peers and with other adults. The results suggest that child care children orient somewhat more strongly to peers and somewhat less strongly to adults than their home-reared counterparts (Belsky, 1984; Belsky and Steinberg, 1978; Clarke-Stewart and Fein, 1983; Rutter, 1981b). This peer orientation appears to have positive as well as negative correlates. Positively, studies indicate greater complexity and reciprocal perceptiveness in the peer interactions of children in child care. In sum- marizing this evidence, Clarke-Stewart and Fein (1983:959) conclude that children with experience in early childhood programs "have been observed to be more popular, . . . to form relationships with other children more often ... and more positively or agreeably...." Negatively, "a number of studies have documented the tendency of children in early childhood programs to be more antisocial with peers" (Clarke-Stewart and Fein, 1983:959~. It is important to note that the measures used in studies do not unambiguously indicate whether the more frequent peer conflict reflects hostile or angry behavior in a clinically problematic range, heightened aggression within a normal range, or simply positive assertiveness. One interpretation of these findings is that "with greater peer exposure comes greater peer interaction,

62 WHO CARES FOR AMERICA'S CHILDREN? which is more likely to be both positive and negative in quality" (Belsky, 1984:13). Several studies have reported children in child care to be less coop- erative with adults (Haskins, 1985; Rubenstein and Howes, 1983; Schwarz et al., 1974~. Schwarz and colleagues (1974), for example, found child care experiences to be associated with greater physical and verbal conflict among preschoolers, as well as with less cooperation with adults. Belsky (1984), in reviewing the findings on child care and cooperation with adults, raises the possibility that problems may be a reflection of particular ex- periences of children in particular child care settings: that is, the effects may be program specific. He notes that while the Swedish research does show more advanced peer relations among children in child care, there is no indication in this research of differences in children's cooperation with adults. Such differences, then, may be a result of specific experiences in child care, rather than an inevitable result of child care participation. Social Competence Clarke-Stewart and Fein (1983) conclude that, in addition to differ- ences in the nature of their relationships with adults and with peers, children in child care show differences in more general social attributes or charac- teristics. In particular, they found the evidence to indicate that children in child care show greater social competence. Thus, for example, in work by Clarke-Stewart (summarized in Clarke-Stewart and Fein [1983~), child care children scored higher than home-reared children on a rating of social competence that encompassed indices of awareness of social norms, appro- priate independence, friendliness, responsiveness, and social confidence. Other studies reviewed by Clarke-Stewart and Fein show differences in the dimensions of social cognition (social problem solving, perspective taking, understanding of emotional labels and sex roles) and of behavior in social situations (self-confidence, self-sufficiency, assertiveness, tendency to be outgoing, helpfulness). In interpreting these findings, Clarke-Stewart and Fein hypothesize that children develop greater social competence in part from the skills required to interact with a range of different peers. There have been few attempts to evaluate the overall socioemotional adjustment of children in child care as opposed to home-reared children. In particular, the use of clinical measures of adjustment, of known psycho- metric properties, has been and continues to be rare in studies of children in child care programs. Therefore, it is difficult to say whether differences observed between the social behaviors and attributes of children in child care and home-reared children resect variations within the normal range or whether from a clinical perspective child care children show indications of more or less adequate overall adjustment. Similarly, research to date

THE EFFECTS OF CHILD CARE 63 has not addressed the possibility that participation in child care settings has distinct implications for the socioemotional development of minority group children. For example, are child care programs that incorporate multicul- tural perspectives associated with more positive cultural identification? Do such program emphases have implications for other aspects of socioemo- tional development in minority group children or for their later adaptation to elementary school? Thus, while there are indications of greater social competence among children in child care, the evidence available to date regarding the overall socioemotional development of children in child care is extremely limited. Methodological Issues The first wave of child care research is characterized by a group comparison strategy: the development of children in child care is compared with that of home-reared children. As child care research has progressed, there has been growing awareness of methodological issues inherent in this approach. This awareness has led both to methodological refinement in studies continuing to use a group contrast approach and to the emergence of a second wave of studies using a different approach. One major methodological issue in the first wave of research is the pos- sibility that group differences are not rooted in the child care experiences, but rather reflect other ongoing differences among child care children and home-reared children and their families (Belsky, 1984~. Such differences may be relevant to some "outcomes." Perhaps, for example, parents who enroll their children in child care are in part responding to an already stronger motivation in their children to interact with peers rather than adults. Perhaps there are differences in the nature of parent-child inter- actions between these groups. Researchers using the two-group approach have called for increased use of random assignment to care settings and the use of home-reared control groups drawn from child care waiting lists, in order to control for possible preexisting tendencies (Cochran, 1977; Lamb, Hwang, Bookstein et al., 1988), and for examination of behavior in child care children and their families prior to entry into child care (Lamb, Hwang, Bookstein et al., 1988; Roopnarine and Lamb, 1978~. Another methodological issue in these studies is that child care samples have been drawn primarily from high-quality, often university-based, model programs. Although these studies can show, for example, whether child care under optimal circumstances involves alteration in the mother-child relationship, they cannot show whether differences occur for the majority of children who attend community-based child care. The increasing inclusion of community-based programs in research using the group comparison

64 WHO CARES FOR AMERICA'S CHILDREN? approach has forced an awareness of the wide range of experiences of children in child care. As we have noted, the large majority of studies using a group compar- ison approach have focused on center care. Thus, there are relatively few data pertaining to the far more widespread family day care. Also, there is very little information on whether participation in child care, particu- larly community-based care, has differential implications, either salutary or stressful, for children from minority cultural, ethnic, and racial groups. And, also as noted, the first-wave studies to date are most often cross-sectional rather than longitudinal and thus fail to yield a picture of differential impli- cations of child care participation (beyond the infancy period) in relation to child age. Finally, researchers have also noted that the group comparison strategy generally fails to tie findings to particular processes or experiences. Child care in the United States seems to enhance both positive and negative behaviors with peers, but what specifically are the features that do so? Are such effects related to caregiver emphasis on guidance in social interactions? other specific qualitative features of the program, such as group size or ratio? The first wave of child care research points to the need both to search for associations between child outcomes and particular features of the child care environment and for refinements in studies using the group comparison approach. Summary One can conclude with some certainty from the first wave of research that child care participation is not inevitably or pervasively harmful to children's development. Indeed, in certain respects, children benefit from experiences in child care. Beyond this broad statement, more detailed conclusions from the first wave of child care research need further scrutiny as methodological refinements become more widespread among studies using the group comparison strategy, notably as studies use sample selection or assignment procedures to control for self-selection and as a wider range of community-based child care settings, including family day care settings, is included in research. At present, however, some conclusions about development among child care participants are possible. In the area of cognitive development, there is no evidence that child care participation has negative effects among middle-class children. Fur- thermore, high-quali~,r cognitive enrichment child care programs have pos- itive implications for intellectual development among low-income children at risk for declining IQ scores. In the area of socioemotional development, the evidence points to a pattern of greater overall social competence in children with child care

THE EFFECTS OF CHILD CARE 65 experience. Children in child care show a pattern of peer interactions that is richer and more complex, but also characterized by more conflict. Children in child care tend to show a shift in social orientation toward peers and away from adults. For children beginning full-time child care in the first year of life (though not for those starting later or less than full time), there are differences in the pattern of their attachment to their mothers. This finding is open to a range of interpretations, however, that will need to be resolved through further research. THE SECOND WAVE OF CHILD CARE RESEARCH: VARIATION IN CHILD CARE QUALITY AND CHILDREN'S DEVELOPMENT As child care research moved beyond model programs to include community-based family day care and center care, it became increasingly clear that child care programs and arrangements are extremely heteroge- neous. They vary from minimally structured and custodial environments to highly structured and enriched environments. The actual teacher/child ratio for 3- or 4-year-old children in centers ranges from 1 teacher per 5 to 1 per 24 (Vandell and Powers, 1983~. Ibys and educational materials can be abundant and in good condition or limited and ragged. Caregiving staff can have college degrees in child development or have no college education or training pertinent to children. There may be low rates of turnover and good continuity of relationships between particular caregivers and children or high rates of staff turnover and poor continuity of relationships. Family day care providers may be isolated from resources and support or part of a supportive network Directors may make use of or generate resources in the community for preventive mental health work with children and families or may not do so. Parents may feel that they collaborate with their child care providers in the care of their children or they may feel criticized and excluded. Does this variation in child care quality have implications for children's development? The second wave of research asks: Does qual- ity of care have an influence on children's development while they are in care? Are there any implications of child care quality that persist into the elementary school years? The research uses three approaches to measuring quality. In many studies, a global or summary measure is used. Researchers distinguish between high- and low-quality care (or high, medium, and low) based on a composite picture of such factors as staff/child ratios, caregiver training, organization of space, and daily routine. A widely used composite measure is the Harms and Clifford (1980) Early Childhood Environment Rating Scale. The second approach is to focus on individual components of overall quality in relation to outcomes. In such an approach, a specific

66 WHO CARES FOR AMERICA'S CHILDREN? component of quality, such as staff/child ratio or group size, is examined in relation to development. The third approach is to examine not the physical or structural features of child care related to quality (e.g., group size, staff/child ratio), but rather to define quality in terms of children's experiences in care. Thus, caregiver verbal behavior or empathic behavior may be related to child development. In contrast to earlier studies, the strategy in this second wave of re- search has been largely naturalistic in that it involves the study of correlates of quality variation as it actually occurs in community-based (rather than model) child care. ~ anticipate the discussion of methodological issues for this wave of research (as well as of our assessment of the needs for future research in Chapter 9), an understanding of quality would be strengthened if the naturalistic approaches to quality were more often complemented with studies involving random assignment and manipulation of quality variables. Although most studies do control for key family background variables, the possibility remains that ongoing characteristics of the families or children themselves may affect both their placement in care of varying quality and child outcomes. Again, to anticipate, the third wave of child care research is beginning to grapple with these issues. Because the second wave of research is still very much in progress, the panel commissioned a detailed review of the relevant findings. The following summary draws extensively on that review (Zaslow, 1988) and on one by Phillips and Howes (1987~. Our discussion here centers on the issue of whether there is evidence that quality of care (as defined in any of the three ways indicated above) matters for contemporaneous or later development. By contrast, our discussion of quality of care in Chapter 4 attempts to specify which of the structural dimensions of quality have the strongest associations with children's development. Quality and Contemporaneous Development Overall, quality of care has been found to be associated with children's cognitive as well as social development when developmental status is as- sessed at the same time as quality of child care. These findings hold for samples that are diverse both as to family background and type of care received. Cognitive Development Analyzing the cognitive and language data from a study of center care in Bermuda, McCartney (1984:251) concluded that, overall, center quality "appears to have a profound effect on language development." A summary

THE EFFECTS OF CHILD CARE 67 measure of quality significantly predicted children's scores on several ac- cepted assessments of language skills, including the revised Peabody Picture Vocabulary Test (PPVT), the Preschool Language Assessment Instrument, the Adaptive Language Inventory, and, for a subsample, ratings of free speech in a communication task Furthermore, the total number of care- giver utterances to children predicted children's scores on the Adaptive Language Inventory and free-speech samples, whereas conversations initi- ated by children with their peers were a negative predictor of three of the four language measures. In the National Day Care Study of center care in the United States (Ruopp et al., 1979), children's change in scores from fall to spring on the Preschool Inventory (PSI), a school readiness test, and the PPVT, were related to center group size, teacher qualifications, and center orientation. In smaller groups, children made greater gains on both measures. In centers in which caregivers had child-related training or education, children made greater gains on the PSI. Finally, children showed greater gains on test scores in centers where staff cited cognitive development as a goal and where the focus was on individual development rather than group experience. As in the study by McCartney (1984), cognitive development in chil- dren in the National Day Care Study was related to observed caregiver behaviors. In centers with lower gains in PSI scores and thus less advanced development, caregivers showed less individual attention to children, en- gaged children in more open-ended and fewer structured activities, and interacted with children more often in large than in medium-sized groups. Greater gains In scores on the PPVT were related to interactions occurring with individual children and with medium-sized groups, with more teacher management of activities, and more social interaction with children. Findings reported by Goelman and Pence (1987a) extend the cogni- tive findings to family day care settings. Studying care quality in Victoria, Canada, they found quality of care to be much more variable in family day care than in center care, and a `'much more potent predictor of chil- dren's language development than quality in centers" (Goelman and Pence, 1987a:9!9~. Total quality scores predicted children's PPVT language scores as well as their scores on the Expressive One-Word Picture Vocabulary Test. Social Development Quality of care is also predictive of children's concurrent social de- velopment from toddlerhood through the preschool years. For example, Anderson and colleagues (1981) noted that the behavior of 2 1/2-year-olds in a laboratory observation differed according to the level of involvement

68 WHO CARES FOR AAlERICA'S CHILDREN? Of the center caregivers with the children. With highly involved caregivers, children showed behavior suggestive of secure attachment: more initial exploration in an unfamiliar room; more physical, visual, and vocal contact with the caregiver; and selective orientation to the caregiver rather than a stranger after a period of time alone in the laboratory playroom. Howes and Olenick (1986) contrasted the compliance and self-regula- tory behaviors of toddlers in high- and low-quality center care. Observa- tions indicated that children in the low-quality settings were less compliant and more resistant. In laboratory observations, children from low-quality centers were less likely to regulate their behavior in a situation requiring restraint. McCartney and colleagues (1982) found ratings of center quality to be related to preschoolers' social as well as cognitive development. In centers in Bermuda with lower overall quality, children were given lower ratings on sociability and considerateness, using the Classroom Behavior In- ventory. Children in centers with less adult talk to individual children were found to be less adult oriented, and the language environment of centers also predicted ratings of child considerateness. Extending the findings to family day care, Clarke-Stewart (1987) reported less optimal social develop- ment in children whose caregivers less often engaged them in conversation and less often touched them, read to them, or gave them directions. A study carried out in Sweden (Lamb, Hwang, Broberg, and Bookstein, 1988), however, does not support the prediction that high-quality care is associated with greater social or personality maturity. In that study, quality of care was related in inconsistent and contradictory ways with outcome measures. For example, children who were observed to be more sociable both with peers and unfamiliar adults were in out-of-home child care (both center and family day care) that was rated lower in both positive and negative events occurring in child care as observed using the Belsky and Walker checklist. Similarly, regarding a measure of personality maturity, quality of out-of-home care was significantly predictive, but not in the expected direction: out-of-home care of lower quality (as measured by the Belsky and WaLker checklist) was associated with greater maturity. Lamb and colleagues urge caution in generalizing to child care in the United States from these findings. They note, in particular, that all out-of-home care in Sweden is of exceptionally high quality by comparison with that in the United States. They raise the possibility that the limited quality variation tapped by their measures is not sufficient for an assessment of the implications of care quality and underlies the contradictory findings. They conclude that consistently high-quality care in Sweden "makes this culture a poor choice for research emphasizing the quality of out-of-home care" (Lamb, Hwang, Broberg, and Bookstein, 1988:39~. Interestingly, social skills as well as personality maturity in that study were positively predicted by time spent in out-of-home care. Thus, more time spent in (uniformly

THE EFFECTS OF CHILD CARE 69 high-quality) Swedish family or center day care positively predicted social development, whereas variation within this limited quality range was not a predictor of social development in a consistent manner. Overall, then, the findings for social development, as for cognitive development, support the prediction that care quality is related to measures of development. Longitudinal Correlates of Child Care Quality Five studies have been carried out in the United States that relate quality of child care at one age to later development. Two of the studies focus on the quality of care during toddlerhood and predict development at age 3 (Carew, 1980; Golden et al., 1978~;2 the other three studies examine quality of care during the preschool years in relation to development in kindergarten or later (Howes, 1988; Howes, in press, a; Vandell et al., 1988~. Each of these studies supports the hypothesis that quality of care has continuing effects. The findings again pertain to family day care as well as center care and to samples of children from differing family backgrounds. Because of their particular relevance for policy, we note especially the findings extending beyond the preschool years. In a prospective study by Howes (1988), children's social and cognitive development was assessed at the end of first grade in a high-quality model elementary school in light of the quality and stability of the children's previous child care experiences. The sample was diverse both in ethnicity and socioeconomic status: the ethnicity and socioeconomic status (SES) of children in the school were selected to match the distributions in the U.S. population. The children had attended 81 different center care and family day care settings. Assessment of the children's development occurred after 3 years in the lab school and encompassed teacher ratings of the child's academic progress and school skills, and parent ratings of behavior problems. With family characteristics controlled, higher quality of earlier child care was predictive of better academic progress and school skills and fewer behavior problems in boys and of better school skills as well as fewer behavior problems in girls. Howes (in press, a) has also reported on analyses from a further study examining age of entry (before or after first birthday) and qual- ity of child care (high or low) in relation to social and cognitive devel 21he studies lay Carew (1980) and Golden and colleagues (1978) are good examples of how waves of day care research have not been entirely separate, but rather overlap. Both studies present results pertinent both to the first wave (group comparisons) and to the second (qualityvariation).

70 WHO CARES FOR AMERICA'S CHILDREN? opment in the toddler and preschool periods and in kindergarten. Quality of child care was predictive of later social outcomes but not of cognitive outcomes. Looking particularly at outcomes during kindergarten, lower quality child care predicted more child hostility and less task orientation as rated by teachers. Children who entered care before their first birthdays and experienced poor-quality care received less positive teacher ratings on distractibility and considerateness in kindergarten. Vandell and colleagues (1988) found that the quality of center care (high or low) that children received at age 4 affected their observed and rated social behavior in three-way peer interactions at age 8. Controlling for social class, higher quality care at age 4 for this white middle-class sample significantly predicted friendlier peer interactions, more positive affect, greater social competence, and better conflict negotiations. Higher quality care at age 4 was negatively correlated with unfriendly interactions, solitary play, and designations of the child as shy. This study also found significant relations between children's observed activities in child care at age 4 (as opposed to overall quality of care at age 4) and their social functioning at age 8. For example, more positive interaction with caregivers at age 4 was related to ratings of the child at age 8 as more socially competent, peer accepted, empathic, and capable of negotiating convicts and frustration. The longitudinal evidence now extends only through the early years of elementary school; it is as yet limited to a small set of studies; and it has not yet eliminated the possibility that further variables may explain the correlations between quality of care and child development. Nevertheless, it is consistent in finding that the quality of center and family day care that children experience in the preschool years is associated with measures of later development. Methodological Issues An important strength of the research on child care quality is its ability to go beyond the model day care programs and more closely describe child care as it actually is experienced by the majority of U.S. children. These studies encompass samples that vary by ethnicity and socioeconomic status, and their findings pertain to family day care as well as center care. As a result, the research on quality clearly permits generalizations beyond white middle-class children in university programs. The use of longitudinal research strategies also reflects an important strength in this wave of research. A problem in this wave of research, however, pertains to the way in which quality is measured. From a policy perspective, the most useful way to assess quality is through specific, potentially regulatable, program characteristics such as group size, caregiver/child ratio, caregiver training,

THE EFFECTS OF CHILD CARE 71 and educational material available. Many of the studies of quality rely instead on global or summary measures. It is impossible to separate the particular program features to determine which are most strongly or causally related to children's development. Thus, the existing research on quality often addresses the broad questions Does quality matter to both immediate and longer term development? rather than the specific question of which aspects of quality matter most (see Chapter 4~. In a similar vein, the research on quality is limited in its usefulness in the policy arena in that it has not, as yet, considered effect sizes. For example, the magnitude of improvements on particular child outcomes cannot be associated with specific increments in quality. As the research has proceeded, a further methodological issue has emerged, that of the relationship between measures of quality and family characteristics. Higher quality care has been found to be associated with family economic and psychological characteristics in a number of studies. Just as the methodological issues that emerged in the first wave of child care research gave rise to the second wave, this key issue is sparking the emergence of a third wave of studies. How and why are family and child care measures linked? Summary The second wave of child care research strongly supports a key con- clusion: child care quality is important to children's development. The strength of the second-wave research to date is that it is very broadly based. The linkage between child care quality and children's development has been documented using a variety of approaches to define quality; samples of varying socioeconomic status; both family day care and center day care settings; and cognitive as well as socioemotional measures of children's development. Furthermore, there are now indications that quality of care in the preschool years continues to have implications for children's development into the early school years. Further methodological progress in the second wave of child care research can be expected in several important areas: wider use of re- search strategies involving manipulation of selected quality dimensions; more efforts at disentangling the component features of quality to examine their relative contributions; and attempts to determine the magnitude of improvement in children's development associated with measured improve- ments in quality. A further methodological issue is already the focus of substantial interest, that of the association between family characteristics and the quality of care a child receives.

72 WHO CARES FOR AMERICA'S CHILDREN? TOWARD A THIRD WAVE OF CHILD CARE RESEARCH: THE LINKAGES BETWEEN FAMILY AND CHILD CARTE ENVIRONMENTS Two types of evidence indicate that family and child care environments are related: findings showing linkages between child care quality and family SES and findings showing associations between child care quality and family social and psychological characteristics. Child Care Quality and Family Social and Psychological Characteristics Not all studies relating SES variables (e.g., parental education, occupa- tion, income) and quality of care report the two to be significantly related (see, e.g., Howes, 1983; Howes and Olenick, 1986; McCartney et al., 1982~. Those that do report a relationship indicate that higher SES is associated with better quality care (e.g., Anderson et al., 1981; Goelman and Pence, 1987b; Holloway and Reichhart-Erickson, 1988; Kontos and Fiene, 1987) or that in lower income samples very low SES is associated with higher quality care, most likely because of the availability of government subsi- dies (Ruopp et al., 1979~. Thus, Anderson and colleagues (1981) found middle-class parents of children in "high physical quality" centers to be better educated, and Goelman and Pence (1987a) found that children from low-resource families (single mothers with little education, low-status occu- pation, and low incomes) disproportionately enrolled in low-quality centers. Yet, in the National Day Care Study (Ruopp et al., 1979), in centers re- ceiving some federal funds, children in classes with better staff/child ratios tended to be from poorer neighborhoods, to have less-educated mothers, and to come from single-parent families. In general, it appears that in the absence of government subsidies, higher quality child care and higher SES are correlated. Recent studies go beyond socioeconomic factors to point to differences in family values and behaviors that are associated with differences in child care quality. For example, Howes and Olenick (1986) report that families using low-quality center care had higher scores on a measure of "complex- ity" (parents live apart, work requires travel, long work hours, weekend work split shift), a variable that can be interpreted to mean family stress. Furthermore, parents of children in high-quality child care were found to be more involved and invested in child compliance. Howes and Stewart (1987) found that families that could be characterized as "nurturing and supported" by a social network had children in higher quality care; more "restrictive and stressed" families had children in lower quality care. In the study of center care in Bermuda, two family background variables were

THE EFFECTS OF CHILD CARE 73 found to be significantly correlated with center quality: the value the family placed on social skills (as measured by parental reports) was positively related to the quality of the child care arrangement; the value the family placed on conformity (again measured through parental reports) was neg- atively correlated with quality (McCartney, 1984; Phillips et al., 1987~. A partial replication within the United States for this result is reported by Kontos and Fiene (1987~: center quality in the state of Pennsylvania was positively associated with the value the family placed on prosocial behavior. Howes (in press, by, in an excellent discussion of the issues, notes that family variables and the quality of care may have mutual influences over a period of time. A stressed family, for example, may not be able to persevere in a search for higher quality care and may place a child in a lower quality setting. Experiences in such a setting then influence the child's development (perhaps, for example, with regard to compliance), which in turn may increase stress levels in the home. Such patterns of mutual influence between school and family have been identified in older children, especially boys (Patterson, 1986~. The Link Between Child Care Quality and Family Variables Given the interrelated nature of family and child care quality measures, recent research has asked whether the quality of care has an impact separate from family economic and psychological variables. Evidence that family variables and the quality of care, separately, contribute to development is of two kinds: correlational studies in which care quality continues to predict child development with family variables controlled, and research designs involving random assignment to different child care situations. Studies using correlational designs have consistently concluded that family and quality of care variables are important contributors. Howes and Stewart (1987), for example, examined the role of family character- istics (factor scores describing families as "nurturing and supported" or "restrictive and stressed") and quality of child care in predicting the level of children's play with objects, peers, and adults. Each set of variables significantly predicted level of play with the other set of variables con- trolled. McCartney (1984) concluded that the quality of center care was as predictive of children's language skills as family background variables. As research using the strategy of controlling for correlated family variables progresses, one can anticipate the assessment of a wider range of family variables and the use of direct observation of family processes, rather than reliance on self-report measures. An excellent recent example of progress in both respects can be seen in the work of Owen and Henderson (1989~. In this study, child care quality at age 4 was found to be related to several measures of mother-child and father-child interaction as observed

74 WHO CARES FOR AMERICA'S CHILDREN? at 12 months: children in higher quality care at age 4 had shown less negative affect at 12 months and had parents who had been rated as more sensitive and positive. However, even with these early observational measures controlled, children in higher quality care at age 4 were observed to show more advanced social skills. A study by McCartney and colleagues (1985) goes beyond a correla- tional design to ask what happens to children of lower socioeconomic status, compared with more advantaged children, when they are in center care of higher overall quality. In this study, the social and cognitive development of children attending a government-run center for low-income children in Bermuda was compared with the development of children attending eight private child care centers on the island. The children at the government centers had mothers not only with lower occupational status, but also with lower verbal IQ scores than mothers in the comparison group. At the same time, the overall quality of the government center was the highest of the centers as measured by the Early Childhood Environment Rating Scale. Analysis of the developmental status of the children in the government center with children of the same age attending all other child care programs indicated higher scores for those in the government center on measures of language development, intelligence, and task orientation, as well as on indices of sociability and considerateness. The authors comment that the findings "are especially convincing because the comparison group consists of children of higher SES . . . " (McCartney et al., 1985:251~. Studies involving random assignment of children to groups varying as to quality of care provide the most rigorous examination of whether quality has an impact independent of family factors. Experimental designs, which randomly assign children to differing child care situations, are a widely used strategy in studies of early intervention for children from disadvantaged families. In a recent review of that evidence, Bryant and Ramey (1987) identified assessments of 17 early intervention programs that involved random assignment to intervention or no-intervention groups. They restricted their attention to studies involving such designs in light of concerns about the failure, in other studies, "to ensure adequately the initial equivalence of educationally treated and untreated (control) groups" (Bryant and Ramey, 1987:35~. The crucial finding from that review is that the program benefits, as measured by IQ scores, were most closely related to the intensity of contact (amount and breadth) a child had with the intervention program. This pattern, identified by looking across studies, has also been found within a study that systematically varied children's extent of contact with an intervention program (Horacek et al., 1987~. This dose-response relationship in experimentally oriented research permits us

THE EFFECTS OF CHILD CARE 75 to conclude with confidence that programs involving high-quality care have positive implications for children, independent of family background factors. Although designs involving randomization are widely used in early intervention studies, they are as yet rarely used in studies of variation in quality in community-based child care programs. A notable exception illustrates the feasibility and usefulness of such a strategy in studying quality. One substudy within the National Day Care Study (NDCS) (Ruopp et al., 1979) was carried out in the Atlanta public school system and involved random assignment of children, within child care centers, to classrooms varying systematically on quality factors. This substudy confirmed the findings of the overall NDCS that children's growth on cognitive measures from fall to spring was linked most closely to the quality component of group size in child care centers. Studies of variation in child care quality in community-based care appear to be at an early and descriptive phase, documenting naturally occurring variation and its correlates. Such a descriptive phase appears both important and necessary (e.g., it revealed the family-quality associations), and it will be especially important in future work on dimensions of quality to follow the path of the intervention studies and the NDCS in using experimental designs. The conclusion best supported by the existing research is that children who are cared for in both child care and the family are influenced by both. As the research in this new wave progresses, an increasingly clear picture will emerge of how care in one environment can offset or complement care in the other. One example of this more complex conceptualization comes from recent research by Howes and colleagues (1988~. In that study, toddlers with insecure attachments to both mother and caregiver showed the least ability to engage in interactions with caregivers while in child care. Children with insecure attachments to their mothers but secure attachments to caregivers showed behaviors indicating that the relationship with the caregiver was compensatory: "These children appeared more socially competent than the children who failed to form compensatory secure relationships with alternative caregivers (1988:415~." The social behavior of toddlers in child care was thus a reflection of relationships both at home and in the child care setting. As research progresses, consideration of indirect as well as direct effects can also be expected. That is, not only are children directly influenced by both of their care settings, but child care can have implications for parents that in turn influence children. Thus, for example, Parker and colleagues (1987) found that mothers who participated in the supportive activities offered by Head Start reported fewer psychological symptoms, more feel- ings of mastery, and greater satisfaction with the current quality of life at the

76 WHO CARES FOR AMERICA'S CHILDREN? end of a year. The researchers note that such changes in the mothers may well have implications for children: "In addition to the direct effects on children of Head Start, future research should also examine the potentially positive indirect effects on children stemming from the enhanced parental well-being that parents' involvement in Head Start programs produces" (Parker et al., 1987:232~. Similarly, Edwards and colleagues (1987) found participation in a high-quality infant day care program to be associated with changes over time in parent-infant interaction. Parents of center children diverged over time from other parents in such behaviors as playing with, holding, and touching their babies. The authors relate the increases in those behaviors among the parents to the child-centered orientation and behavior of the center caregivers. Indeed, the changes in parental behavior mirrored observed caregiver behaviors. It will be important, then, in future work to examine further the implications for parents of their children's participation in programs of varying quality and to complete the picture by determining the indirect effects of these influences on children. Summary Work reveals that a child's experiences at home and in a child care setting are not separate and unrelated experiences: they are very much linked. Although family and care quality factors are not independent, both sets of factors contribute to children's development: that is, children who experience care both in child care settings and in the home show the influence of both. Thus far, the major focus of the third wave of research has been to document that a child's placement in child care of higher or lower quality in part reflects family psychological and socioeconomic factors. In the absence of subsidies or interventions, families that are more stressed, both psychologically and economically, are more likely to use lower quality care. The United States thus has a group of children in double jeopardy: the children in greatest need of high-quality care to offset stress at home often receive low~ualibr care. Increasingly thorough and rigorous research on the joint contributions of home and child care factors to children's development can be expected. For example, a few studies now identify and control for a wide range of family factors in considering the impact of care quality, but future work on this issue will have to incorporate views of the family-day care linkage that go well beyond the finding that family factors influence choice of care quality. For example, how care quality influences family stress levels and parent-child interaction needs to be examined.

THE EFFECTS OF CHILD CARE 77 CONCLUSIONS The research on child care is evolving, and our evaluation of the evidence points to the continuing need to address issues of methodology. Even as it stands, however, the existing evidence from each of the research stages we have identified provides the basis for broad conclusions: Child care participation is not a form of maternal deprivation. Children can and do form attachment relationships to multiple caregivers, if the number of caregivers is limited, the relationships enduring, and the caregivers are responsive to the individual child. · Child care is not inevitably or pervasively harmful to children's de- velopment. Indeed, the evidence points to aspects of development for which child care is beneficial. The quality of child care in either family day care or center care- is important to children's development, whatever their socioeco- nomic levels and whether one looks at cognitive or socioemotional development. Children from families enduring greater psychological and eco- nomic stress are more likely to be found in lower quality care settings. Thus, there are children in the United States, especially those from low-income families, in double jeopardy from stress both at home and in their care environments. There is no strong basis in our review for urging parents toward or away from enrolling children in child care settings, although we do find unresolved questions concerning full-time care in the first year of life. Rather, our review strongly directs attention to the issue of child care quality and its impact on children's development. Accordingly, the next chapter addresses in greater detail two key questions: What are the components of care quality? What are the thresholds demarcating high- and low-qualib, care on these dimensions? REFERENCES Ainsworth, M.D.S., M. Blehar, E. Waters, and S. Wall 1978 Pattems of Attachment. Hillsdale, NJ.: Erlbaum. Anderson, C.W., RJ. Nagle, W.N Roberts, and J.W. Smith 1981 Attachment to substitute caregivers as a function of center quality and caregiver involvement. Child Development 52:53-61. Barglow, P., B.E. Vaughn, and N. Molitor 1987 Effects of maternal absence due to employment on the quality of infant-mother attachment in a low-risk sample. Child Development 58:945-954. Belsly, J. 1984 INTO waves of day care research: Developmental effects and conditions of

78 WHO CARES FOR AMERICA'S CHILDREN? quality. Pp. 1-34 in R.C. Ainslie, ea., The Child and the Day Care Setting;: Qualitative Vanations and Development. New York: Praeger. 1988 The "effects" of infant daycare reconsidered. Early Childhood Research Quarterly 3:235-272. Belsky, J., and M. Rovine 1988 Nonmaternal care in the first year of life and the security of infant-parent attachment. Child Development 59:157-167. Belsky, J., and L" Steinberg 1978 The erects of daycare. Child Development 49:929-949. Benn, R.K. 1986 Factors promoting secure attachment relationships between employed mothem and their sons. Child Development 57:1224-1231. Bowlby, J. 1951 Matemal Care and Mental Health. Geneva: World Health Organization. 1969 Attachment and Loss, Vot 1, Attachment. New York: Basic Books. 1973 Attachment and Loss, VoL 2, Separation. New York: Basic Books. 1980 Attachment and Loss, VoL 3, Loss. New York: Basic Books. Brazelton, T.B. 1986 Issues for working parents. American Joumal of Orthopsychiatry 56:14-25. Bronfenbrenner, U., and A C. Crouter 1982 Work and family through time and space. Pp. 39-83 in S.B. Kamerman and C.D. Hayes, eds., Families That Work: Children in a Changing World. Panel on Work, Family, and Community, Committee on Child Development Research and Public Policy, Commission on Behavioral and Social Sciences and Education, National Research Council. Washington, D.C.: National Academy Press. Brown, G., and T. Harris 1978 Social Origins of Depression: A Study of Psychiatric DZsorders in Women. London: Tavistock. Bruner, J. 1983 Child's Talk. New York: Norton. Bryant, D.M., and C.T. Ramey 1987 An analysis of the effectiveness of early intervention programs for high-risk children. Pp. 33-78 in M. Guralnick and C. Bennett, eds., The Effectiveness of Early Intervention for At-Risk and Handicapped Children. New York: Academic Press. Campos, J.J., K.C. Barrett, M.E. Lamb, H.H. Goldsmith, and C. Stenberg 1983 Socioemotional development. Pp. 783-915 in P.H. Mussen, ea., Handbook of Child Psychology, VoL 2. New York: Wiley. Carew, J. 1980 Experience and the development of intelligence in young children at home and in day care. Monographs of the Society for Research in Child Development 45~6-73:Serial No. 187. Chibucos, T., and P. Kail 1981 Longitudinal examination of father-infant interaction and infant-father interac- tion. Merrill-Palmer Quarterly 27:81-96. Clarke-Stewart, K.A. 1987 Predicting child development from child care forms and features: The Chicago study. Pp. 21-42 in D.A Phillips, ea., Quality in Child Care: What Does Research Tell Us? Washington, D.C.: National Association for the Education of Young Children. 1989 Infant day care: Maligned or malignant? American Psychologist 44:266-274.

THE EFFECTS OF CHILD CARE 79 Clarke-Stewart, K.A. , and G. G. Fein 1983 Early childhood programs. Pp. 917-999 in PH. Mussen, ea., Handbook of Child Psychology, HoL 2. New York: Wiley. Cochran, M. 1977 A comparison of group and family child-rearing patterns in Sweden. Child Development 48:702-707. Crockenberg, S. 1981 Infant irritability, mother responsiveness, and social support influences on the security of infant-mother attachment. Child Development 52:857-865. Darlington, R.B., J.M. Royce, AS. Snipper, H.W. Murray, and I. Lazar 1980 Preschool programs and the later school competence of children from low income families. Science 208:202-204. Doyle, A. 1975 Infant development in day care. Developmental Psychology 11:655-666. Dunn, J. 1983 Sibling relationships in early childhood. Child Development 54:787-811. 1988 The Beginnings of Social Understanding. Cambridge, Mass.: Harvard University Press. Dyer, E.D. 1963 Parenthood as crisis: A re-study. Marriage and Family Living 25:196-2~1. Edwards, C.P., M.E. Logue, S.R. Loehr, and S.B. Roth 1987 The effects of day care participation on parent-infant interaction at home. American Joumal of Orthopsychiatry 57:116-119. Farber, E.A., and B. Egeland 1982 Developmental consequences of out-of-home care for infants in a low-income population. Pp. 102-125 in E.F. Zigler and E.W. Gordon, eds., Day Care: Scientific and Social Policy Issues. Boston: Auburn House. Freud, A., and D.T. Burlingham 1944 Infants Without Families; The Case for and Against ResidentialNurseries. New York: International Universities Press. 1973 Infants Without Families; Reports on the Hampstead Nurseries, 1939-1945. New York: International Universities Press. Goelman, H., and A R. Pence 1987a Effects of child care, family, and individual characteristics on children's lan- guage development: The Victoria Day Care Research Project. Pp. 89-104 in D.A. Phillips, ea., Quality in Child Care: What Does Research Tell Us? Washington, D.C.: National Association for the Education of Young Children. 1987b Some aspects of the relationship between family structure and child language in three types of day care. Pp. 129-146 in D. Peters and S. Kontos, eds., Annual Advances in Applied Developmental Psychology, VoL 2. Norwood, NJ.: Ablex Publishing Corp. Golden, M., L Rosenbluth, M. T. Grossi, H.J. Policare, H. Freeman, Jr. , an d E.M . Brownlee 1978 The New York Cay Infant Day Care Study. New York: Medical and Health Research Association of New York City. Goldfarb, W. 1943 The erects of early institutional care on adolescent personality. Joumal of ExpeKanental Education 12:106-129. Grossman, K.E., K Grossman, F. Huber, and U. Wartner 1981 German children's behavior towards their mothers at 12 months and their fathers at 18 months in Ainsworth's Strange Situation. Ir~temational Joumal of Behavioral Development 4:157-181.

80 WHO CARES FOR AMERICA'S CHILDREN? Harms, T., and R.M. Clifford 1980 Early Childhood Environment Rating; Scale. New York: Teachers College Press. Haskins, R. 1985 Public school aggression among children with varying day-eare experience. Child Development 56:689-703. 1989 Beyond metaphor: lathe efficacy of early childhood education. American Psy- cholog~st 44:274-282. Hobbs, D.F., and S.P. Cole 1976 Transition to parenthood: A decade replication. Joumal of Marriage and the Family 38:723-731. Hoek, E., K Christman, and M. Hoek 1980 Factors associated with decisions about return to work in mothers of infants. Developmental Psychology 16:535-536. Hoek, E., M.T. Gnezda, and S.L. McBride 1984 Mothers of infants: Attitudes toward employment and motherhood following birth of the first child. Joumal of Mamage and the Family 46:425-431. Hoffman, L^W. 1979 Maternal employment: 1979. American Psychologist 34:859-865. Holloway, S.D., and M. Reichhart-Erickson 1988 The relationship of day-eare quality to children's free play behavior and social problem solving skills. Early Childhood Research Quarterly 3:39-54. Horaeek, H.J., C.T. Ramey, F.A. Campbell, KP. Hoffmann, and R.H. Fletcher 1987 Predicting school failure and assessing early intervention with high-risk children. Joumal of the American Academy of Child and Adolescent Psychiatry 26:758-763. Howes, C. 1983 Caregiver behavior and conditions of earegiving. Joumal of Applied Develop mental Psychology 4:99-107. 1988 Relations between early child care and schooling. Developmental Psychology 24:53-57. in press, a Can age of entry and the quality of infant child care predict behaviors in kindergarten? Developmental Psychology. in press, b Current research in early day care: A review. In S. Chehrazi, ea., Balancing Working and Parenting Psychological and Developmental Implications of Day Care. New York: American Psychiatric Press. Howes, C., and M. Oleniek 1986 Family and child influences on toddlers' compliance. Child Development 57:202- 216. Howes, C., and P. Stewart 1987 Child's play with adults, toys, and peers: An examination of family and child care influences. Developmental Psychology 23:423~30. Howes, C., C. Rodning, D.C. Galluzzo, and L. Myers 1988 Attachment and child care: Relationships with mother and earegiver. Early Childhood Research Quarterly 3:403-416. Kontos, S., and R. Gene 1987 Child care quality, compliance with regulations, and children's development: The Pennsylvania study. Pp. 57-80 in D.A. Phillips, ea., Quality in Child Care: What Does Research Tell Us? Washington, D.C.: National Association for the Education of Young Children. Lamb, M.E., C.-P. Hwang, F.L" Bookstein, ~ Broberg, G. Hull, and M. Frodi 1988 Determinants of social competence in Swedish preschoolers. Developmental Psychology 24:58-70.

THE EFFECTS OF CHILD CARE 81 Lamb, M.E., C.-P. Hwang, A. Broberg, and F.L. Bookstein 1988 The erects of out-of-home care on the development of social competence in Sweden: A longitudinal study. Early Childhood Research Quarterly 3:379-402. Lazar, I., R.B. Darlington, H. Murray, J. Royce, and A. Snipper 1982 Lasting effects of early education: A report of the Consortium for Longitudinal Studies. Monographs of the Society for Research in Child Development 47~2- 3~:Serial No. 195. Lee, V.E., J. Brooks-Gunn, and E. Schnur 1988 Does Head Start work? A 1-year follow-up comparison of disadvantaged children attending Head Start, no preschool, and other preschool programs. Developmental Psychology 24:210-222. LeMasters, E.E. 1957 Parenthood as crisis. Marriage and Family Irving 19:352-355. Lock, A., ed. 1~78 Aciron, Gesture, and Symbol. New York: Academic Press. Matas, L., R. Arend, and L.N Sroufe 1978 Continuity of adaptation in the second year. The relationship between quality of attachment and later competence. Child Development 49:547-556. McCartney, K. 1984 Effect of quality of day care environment on children's language development. Developmental Psychology 20:244-260. McCartney, K., S. Scarr, D. Phillips, and S. Grajek 1985 Day care as intervention: Comparisons of varying quality programs. Joumal of Applied Developmental Psychology 6:247-260. McCartney, K., S. Scarr, D. Phillips, S. Grajek, and J.C. Schwarz 1982 Environmental differences among day-care centers and their effects on children's development. Pp. 126-151 in E.F. Zigler and E.W. Gordon, eds., Daycare: Scientific and Social Policy Issues. Boston: Auburn House. Moen, P. 1989 Working Parents: Transformations in Gender Roles and Public Policies in Sweden. Madison: University of Wisconsin Press. Owen, M., and M. Cox 1988 Maternal employment and the transition to parenthood. Pp. 85-119 in NE. GottEried and AW Gottfried, eds., Matemal Employment and Ch~l~en's Devel- opment: Longitudinal Research. New York: Plenum. Owen, M.T., and V.K. Henderson 1989 Relations Between Child Care Qualities and Child Behavior at Age 4: Do Parent/Child Interactions Play a Role? Paper presented at the meetings of the Society for Research in Child Development, Kansas City, April. Parker, F.L, CS. Piotrkowski, and L. Peay 1987 Head Start as social support for mothers: The psychological benefits of involvement. American Journal of Orthopsychiatry 57:220-233. Pastor, D.L. 1981 The quality of mother-infant attachment and its relationship to toddlers' initial sociability with peers. Developmental Psychology 17:326-335. Patterson, G.R. 1986 Performance models for antisocial boys. American Psychologist 41:432-444. Phillips, D. 1988 Quality in Child Care: Definitions at the AL. Mailman Family Foundation, Inc. Paper presented at symposium on dimensions of quality in programs for children. White Plains, New York, June.

82 WHO CARES FOR AMERICA'S CH LDREN? Phillips, D.A., and C. Howes 1987 Indicators of quality in child care: Review of research. Pp. 1-20 in D.A. Phillips, ea., Quality in Child Care: What Does Research Tell Us? Washington, D.C.: National Association for the Education of Young Children. Phillips, D.A., S. Scarr, and K. McCartney 1987 Dimensions and ellects of child care quality: The Bermuda study. Pp. 43-56 in D.A. Phillips, ea., Quality in Child Care: What Does Research Tell Us? Washington, D.C.: National Association for the Education of Young Children. Provence, S., and R.C. Lipton 1962 Infants in Institutions: A Comparison of Their Development With Family-Reared Infants During the First Year of Life. New York: International Universities Press. Ribble, M.A. 1965 The Rifts of Infants: Early Psychological Needs and Weir Satisfaction (2nd end.. New York: Columbia University. Richtem, J.E., and C. Zahn-Waxler 1988 The infant day care controversy: Current status and future directions. Early Childhood Research Quarterly 3:319-336. Robertson, J., and J. Robertson 1971 Young children in brief separations: A fresh look. Psychoanalytic Study of the Child 26:264-315. Roopnarine, J.L, and M.E. Lamb 1978 The effects of day care on attachment and exploratory behavior in a strange situation. Merrill-Palmer Quarterly 24:85-95. Rubenstein, J.L, and C. Howes 1983 Social-emotional development of toddlers in day care: The role of peers and individual differences. In S. Kilmer, ea., Early Education and Day Care, Vol. 3. Greenwich, Conn.: JAI Press. Rubenstein, J.L, C. Howes, and P. Boyle 1981 A two-year follow-up of infants in community-based day care. Joumal of Child Psychology and Psychiatry 22:209-218. Ruopp, R., J. leavers, F. Glantz, and C. Coelen 1~79 Children at the Center: Final Results of the National Day Care Study. Boston: Abt Associates. Rutter, M. 1979 Maternal deprivation, 1972-1978: New findings, new concepts, new approaches. Child Development 50:283-305. 1981a Matemal Deprivation Reassessed. Middlesex, England: Penguin Books. 1981b Social-emotional consequences of daycare for preschool children. American Joumal of Orthopsychiatry 5:4-28. Schwarz, J.C. 1983 Infant day care: Effects at 2, 4, and 8 years. Paper presented at the meeting of the Society for Research in Child Development, Detroit. SchwaIz, J.C., R. G. Strickland, and G. Krolick 1974 Infant day care: Behavioral effects at preschool age. Developmental Psychology 10:502-506. Slaughter, D.T. 1983 Early intervention and its effects on maternal and child development. Mono- g~aphs of the Society for Research in Child Development 48~4~:Serial No. 202. Spitz, R.A 1945 Hospitalism: An inquiry into the genesis of psychiatric conditions in early childhood. Psychoanalytic Study of the Child 1:53-74.

THE EFFECTS OF CHILD CARE 83 Thompson, R.A. 1988 The edects of infant day care through the prism of attachment theory: A critical appraisal. Early Childhood Research Quarterly 3:273-282. Thompson, R.A., M.E. Lamb, and D. Estes 1982 Stability of infant-mother attachment and its relationship to changing life circumstances in an unselected middle class sample. Child Development 53:144- 148. Itevarthen, C 1977 Descriptive analyses of infant communicative behaviour. In H.R. Schaffer, ea., Studies in Mother-Infant Interaction. London: Academic Press. Vandell, D.L, and C.P. Powers 1983 Day care quality and children's free play activities. American Joumal of Orthopsychiatry 53:493-500. Vandell, D.L, V.K Henderson, and K.S. Wilson 1988 A longitudinal study of children with day-care experiences of varying quality. Child Development 59:1286-1292. Vaughn, B.E., K.E. Deane, and E. Waters 1985 The impact of out-of-home care on child-mother attachment quality: Another look at some enduring questions. In I. Bretherton and E. Water, eds., Growing Points of Attachment Theory and Research. Monographs of the Society for Research in Chdd Development 50~1-2~:Serial No. 209. Watem, E., J. Wippman, and L^A. Sroufe 1979 Attachment, positive affect, and competence in the peer group: livo studies in construct validation. Child Development 50:821-829. Wolkind, S. 1974 The components of "affectionless psychopathy" in institutional children. Journal of Child Psychology and Psychiatry 15:215-220. Zaslow, M.J. 1988 The Quality of Care. Paper prepared for the Panel on Child Care Policy, Committee on Child Development Research and Public Policy, Commission on Behavioral and Social Sciences and Education, National Research Council, Washington, D.C. Zaslow, M.J., F.N Pedersen, J.T.D. Suwalsky, and B.A. Rabinovich 1989 Maternal employment and parent-infant interaction at one year. Early Childhood Quarterly 4:459-478. Zaslow, M.J., B.A. Rabinovich, and J.T.D. Suwalsky in press From maternal employment to child outcomes: Preexisting group differences and moderating variables. In J. Lerner and N. Galambos, eds., Matemal Employment in the Child~earing Years. New York: Garland Press.

Next: 4 Quality of Child Care: Perspectives of Research and Professional Practice »
Who Cares for America's Children? Get This Book
×
Buy Hardback | $49.95
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Few issues have aroused more heated public debate than that of day care for children of working parents. Who should be responsible for providing child care—government, employers, schools, communities? What types of care are best?

This volume explores the critical need for a more coherent policy on child care and offers recommendations for the actions needed to develop such a policy.

Who Cares for America's Children? looks at the barriers to developing a national child care policy, evaluates the factors in child care that are most important to children's development, and examines ways of protecting children's physical well-being and fostering their development in child care settings. It also describes the "patchwork quilt" of child care services currently in use in America and the diversity of support programs available, such as referral services.

Child care providers (whether government, employers, commercial for-profit, or not-for-profit), child care specialists, policymakers, researchers, and concerned parents will find this comprehensive volume an invaluable resource on child care in America.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!