Caregivers and teachers form important relationships with the children in their care. Through these relationships, critical behaviors, such as language use with children, emotional tone and warmth, responsiveness and sensitivity, and intentional teaching, all directly affect child development and have longer term effects on later schooling and social-emotional adjustment (NRC, 2001; NRC and IOM, 2000). Evidence is mounting that the behaviors of teachers and caregivers are important; however, determining how to identify, prepare, and support those who practice these desirable skills is more challenging. Researchers and policy makers both have an interest in determining which characteristics (e.g., amount of training, years of experience, possession of a college degree) are associated with the practices that promote healthy child development, as well as how to produce these practices on a large scale across the early childhood care and education (ECCE) workforce, a topic addressed in Chapter 5. This chapter includes presentations and discussion of the research on these important behaviors and characteristics of the workforce, the effects of working conditions of the workforce, as well as the implications of these working conditions for teachers, caregivers, and children.
A great deal of research has focused on the relations between children’s experiences in non-parental care and their short- and long-term
development. Though a complete review of these data was beyond the scope of this workshop, the National Institute of Child Health and Human Development (NICHD) Study of Early Child Care and Youth Development1 is a landmark study of these early experiences. The study included a full range of settings and income groups and serves as a rich source of data regarding the effects of the workforce on child development. Aletha Huston, the Priscilla Pond Flawn Regents Professor of Child Development at the University of Texas at Austin, described some of the findings from this work.
This longitudinal study involves a number of research teams and has followed children in 10 locations since 1991, collecting data on those who were in any type of non-maternal child care for more than 10 hours a week. The researchers observed the child care when the children were ages 6, 15, 24, 36, and 54 months. The observers recorded minute-by-minute documentation of the child’s experiences and used rating scales to assess aspects of the caregiving environment. Researchers also collected data on the caregivers,2 covering issues such as their training and education, beliefs and attitudes, reasons for becoming child care workers, and psychological well-being.
Huston used a model of the relations among caregivers’ characteristics, the features of the caregiving environment, the behavior of the adults and children in the caregiving setting, and outcomes for children to organize her discussion (see Figure 4-1). Researchers have examined these constructs and the relations among them to study interactions and differences that may occur among children of varying ages.
For infants, the researchers used a measure of “positive caregiving quality,” a composite of caregiver sensitivity, positive regard for the child, cognitive stimulation, engagement, and other features, to assess the quality of the adult–child interactions in the child care setting. They observed higher quality when several factors were in place:
Low child-to-adult ratio;
Small group size;
Caregivers with non-traditional child-rearing beliefs;3 and
High-quality physical environment (e.g., amount and types of various materials, health and safety features).
1 See http://www.nichd.nih.gov/research/supported/seccyd/overview.cfm for details about the NICHD.
2 This section of the report uses the term “caregivers” as it was used in the workshop presentation.
3 This characteristic is also sometimes called a child-centered approach, Huston explained, and it refers to the caregiver’s conviction that giving the child opportunities to develop autonomy, express feelings, and make decisions while still setting limits is important.
These factors, Huston explained, had stronger relations to outcomes than did caregivers’ experience, training, and formal education.
For toddlers (ages 15 to 36 months), the importance of the child-to-caregiver ratio for predicting quality declined, and by age 3 the ratio was not related to quality. The degree to which non-traditional beliefs contributed to quality increased in importance with the age of the children, and the importance of caregiver education also increased, though training did not add much predictive value.
Huston also reported on an analysis of caregivers who worked with 2-year-olds, in which the researchers identified five caregiver characteristics as important to outcomes for children (Malerba, 2005):
Education, formal training, and non-traditional beliefs about child rearing;
Years of experience and age;
Conscientiousness about the job and commitment to caring for young children, and low levels of depression;
Finding personal rewards in the job; and
Professionalism and recent training (among family child care providers only).
Of these, caregiver non-traditional beliefs, professionalism, and concerns about caring for children all predicted observed positive engagement with children on the caregivers’ part, but features of the environment (e.g., ratios, healthy practices) were stronger predictors. In child care homes, the caregiver characteristics were weak predictors. Features of the environment (e.g., ratio, having a schedule) were also strong predictors here. These researchers examined the relationships between meeting external quality standards and positive outcomes for children (NICHD, 1999). They determined the extent to which centers met external standards for quality care established by professional or regulatory orga-
nizations, which address features such as ratio, group size, caregiver education level, and caregiver training. The quality criteria do appear to be important. For example, low child-to-adult ratios for children at ages 2 and 3 predict both high levels of positive social behavior and low levels of problem behavior in the children. For 3-year-olds, caregiver education and training predicted school readiness, high language comprehension, and low levels of behavior problems. Analyses that examined the experiences of children at 54 months also found that caregivers’ education, training, and ratios were associated with children’s development of cognitive and social competencies. For children in family child care, the caregiver’s non-authoritarian beliefs and education were both associated with children’s cognitive skills.
Huston summarized the key findings from these analyses:
For the youngest children, the structural features of the environment, particularly the child-to-adult ratios, often outweigh caregiver characteristics.
The relative importance of caregiver characteristics and ratios changes from infancy to preschool settings; ratio is especially important in the first 2 or 3 years. Caregiver beliefs and training are especially important for older children.
Non-traditional, child-centered beliefs about child rearing mediate at least part of the benefits of caregiver education and formal training, although the causal links are not understood. Possibilities are that people who have those beliefs are more likely to seek education, or that education and training can influence those beliefs, for example, or that both are true.
Huston suggested that policy makers should consider children’s ages in defining quality of care, and should develop standards that are appropriate for different age groups. Training might be more effective if it dealt directly with beliefs and attitudes about child rearing, rather than just curriculum and strategies. At the same time, even the best trained caregivers are limited by the structural and environmental constraints of their settings. Policy makers may wish that improved training could compensate for high ratios for the youngest children, but the evidence seems to suggest otherwise. Thus, she observed, “if we are really thinking about attempts to upgrade the workforce, then it involves upgrading the environments in which people work, as well as upgrading their skills.”
One participant noted that the results discussed were not derived from experimental or quasi-experimental designs and that, despite extensive controls for family and child characteristics, they do not support causal inferences. For example, some of the positive association between
a particular teacher or center characteristic and a child outcome could be the result of unmeasured factors that lead parents to place their child in high-quality care, and that would have led to good outcomes for the child even in the absence of high-quality care.
The Impact of Teacher Qualifications
In recent years, a great deal of attention in research and policy has focused on teacher qualifications as a means of identifying those best prepared to deliver high-quality experiences for children. Specifically, research has sought to determine the relationship of teacher degrees and/ or amount or types of training to the quality of their caregiving and teaching, as well as to children’s outcomes. Debate in the field about whether early childhood teachers should be required to hold bachelor’s (B.A.) degrees has been fueled in part by mixed research findings on the topic (Early et al., 2007; Helburn, 1995; Whitebook and Ryan, 2011). Two researchers, Margaret Burchinal, senior scientist at the Frank Porter Graham Child Development Institute at the University of North Carolina, and Steve Barnett, codirector of the National Institute for Early Education Research, who have examined this topic in detail, presented research findings and their views.
High-quality child care experiences are consistently related to better outcomes for children in general, and especially for low-income children, Burchinal observed. Experimental studies have shown that the effects of at least moderate- to high-quality care are evident through adulthood. High-quality care, she explained, involves close teacher–child relationships, frequent sensitive interactions between the child and the teacher, high-quality instruction, and respectful and effective behavior management. High-quality care also involves other features, such as rich physical environment.
Early research suggested that a B.A. degree successfully identified teachers who provided high-quality care (Burchinal et al., 2000; Helburn, 1995; Whitebook et al., 1990). Policy makers responded with requirements that teachers have a B.A., and teaching certificates for early childhood were established accordingly. Head Start instituted a policy requiring that 50 percent of teachers in a center have a B.A., and many state prekindergarten programs had similar policies.
However, Burchinal explained, more recent work (Early et al., 2007; Whitebook and Ryan, 2011) has challenged the findings about the relevance of a B.A. She suggested several possible explanations. Cohort effects could explain the differences between the groups who had B.A. degrees and worked in early childhood at the time of the early studies and those who have that profile now. Differences in the measures and methods
the researchers used is another possible explanation, so Burchinal and her colleagues conducted a secondary analysis of data from seven large preschool projects, including evaluations of prekindergarten programs, longitudinal studies of community programs, and others (Early et al., 2007). They used consistent definitions of teacher education levels (e.g., type of degree and certification), and standard measures of quality and child outcomes (e.g., Early Childhood Environmental Rating Scale [ECERS], Woodcock-Johnson Applied Problems).4 They also used consistent analytic methods to account for missing data and nesting of children in classrooms. Across 27 separate analyses, they found 5 that showed statistically significant effects favoring possession of a B.A., 4 that showed a benefit for possession of any degree, and 2 that showed a benefit for teacher certification.
In Burchinal’s view, the findings show that measures of child care quality or child outcomes were not consistently related to any of the ways of measuring teacher education. The researchers considered possible explanations for these findings. One possibility is that many of the preservice programs for teachers included in the studies were new and their instruction might not have reflected current research regarding effective practices. This hypothesis is supported by a 2008 web-based survey of programs that provide degrees in ECCE. The researchers found that although these programs relied on standards for both coursework and fieldwork, few were actually focusing on developing supportive teacher–child relationships, and few included research in their teaching about practices (Hyson et al., 2008). For example, only 46 percent of the teacher preparation programs cited teachers’ interactions with children as an important focus. In general, the programs were understaffed and heavily reliant on part-time faculty. Thus, the quality of these teacher preparation programs appears to be very uneven. They are typically small and underfunded, and have seen large enrollment increases without commensurate increases in resources to meet the demand.
The good news, Burchinal noted, is that some training programs for existing teachers can be effective at improving quality and child outcomes. A meta-analysis of studies of training programs that focused on direct teaching practices or interactions with children found that programs that were tightly tailored and used manuals to address a specific issue had positive effects (Fukkink and Lont, 2007). Other work, Burchinal added, supports the idea that intensive training programs with clearly defined curriculums and coaching—for example, using defined curriculums, video observations, and onsite mentoring for both entry-level and
4 ECERS is a rating scale for early childhood programs; the Woodcock-Johnson is a tool for measuring cognitive and academic skills.
more experienced staff—can be effective (Clements and Sarama, 2007; Dickinson and Caswell, 2007; Lonigan and Whitehurst, 1998; Pianta et al., 2008; Powell et al., 2010; Wasik et al., 2010).
In summary, Burchinal believes the research does not currently support the B.A. as a critical factor in the quality of early childhood teachers. “Quality can be improved,” she suggested, “when teachers receive carefully selected and implemented professional development, either during pre- or in-service training. We should move beyond a focus on whether the lead teacher has a B.A. to a focus on the content and quality of the higher education program.” Additional research on the specific course requirements or internship practices, approaches to induction of new teachers, coaching, and other strategies, she added, could help pinpoint the best ways to improve the quality of teachers and outcomes for children. She believes that combining effective higher education with onsite induction and coordinated professional development are the most promising avenues for progress.
Barnett provided another perspective on the evidence regarding teacher quality and qualifications. He referenced an article he coauthored with several colleagues, including Burchinal, which addressed key policy questions about the effects of preschool education (Pianta et al., 2009). In that article, the authors assert that the ways that teachers interact with children and their ability to effectively implement appropriate curriculums may matter more to child outcomes than their qualifications, in part because these behaviors “do not appear to be produced in a reliable manner by typical teacher preparation” (Pianta et al., 2009, p. 50). Informed by this work, Barnett offered his own views on an alternative set of questions that should guide future data collection and analysis:
How do children’s learning and development vary with teacher characteristics?
What teacher characteristics are needed to achieve goals for early childhood care and education (e.g., to close 50 or 75 percent of the learning gap for children entering kindergarten)?
Do the effects of teacher characteristics vary with other program features and policies, and with the populations served?
Under what conditions do teacher qualifications make (or not make) a substantive difference?
On the question of the characteristics teachers need to meet desired goals, Barnett noted that the National Research Council (2001) report, Eager to Learn: Educating Our Preschoolers, provided a good definition of what prekindergarten teachers need to know (see Box 4-1) and offered a valuable basis for this kind of research. Different types of research studies
Skills and Knowledge Prekindergarten Teachers Need
• Current understanding of how to put up-to-date knowledge of teaching, learning, and child development into practice—particularly specific foundational knowledge of socio-emotional development and mathematics, science, linguistics, and literature;
• Capacity to provide rich conceptual experiences that promote growth in specific content areas (e.g., mathematics, science) and in broad domains (e.g., language, cognition);
• Effective teaching strategies, including, but not limited to, new knowledge regarding how children learn to read;
• Capacity to identify appropriate content for preschool children;
• Assessment procedures to inform instruction;
• Teaching practices for children who are not fluent in English, come from different cultural/social backgrounds, have disabilities, or differ from the normal range of development; and
• Capacity to work with parents and other family members.
SOURCE: Adapted from NRC, 2001.
can be useful, he observed. Research on effective parenting, studies of program effectiveness, natural variation studies that examine children before and after their early learning experiences, and studies of changes over time resulting from policy shifts (e.g., changes in requirements for teachers) all may contribute useful knowledge.
One recent study provides an overview of research on program characteristics’ effects on children, though Barnett noted that the measures of teacher qualifications are fairly crude because the sampled studies did not all use the same coding procedures (Camilli et al., 2010). This study found positive effects for programs that incorporated intentional teaching (i.e., purposeful and planned activities to achieve specific educational objectives) and also for programs in which children received significant amounts of one-on-one attention. Like Burchinal, Barnett noted that some meta-analyses show small effects for possession of a B.A. (Early et al., 2007; Kelley and Camilli, 2007).
Looking at the available findings another way, Barnett noted that the prekindergarten programs that have produced significant gains for children, particularly in randomized trials, all shared several factors: well-educated teachers (B.A. or higher); adequate compensation (comparable to public school salaries); strong curriculum and professional development; small classes and reasonable teacher-to-child ratios; strong supervi-
sion, monitoring, and review; and both high standards and continuous improvement. Without all of these features, Barnett observed, a program would be unlikely to achieve the same results: “We don’t see one that produced great results—like Abecedarian, Perry Preschool, or Chicago Child-Parent Centers—that did not have all of those characteristics.”
This general finding is important, but Barnett believes existing research has significant limitations. First, no randomized controlled studies of teacher characteristics have been completed. Second, the measures of child learning and development that are used as dependent variables are “narrow and highly imperfect.” He noted that tools such as the Woodcock-Johnson test (see earlier footnote, this section) are limited measures of children’s educational outcomes and are not likely to be good measures of teachers’ contributions to those outcomes. Similarly, rating scales such as the ECERS are useful measures of classroom and program environments, but they are not sophisticated enough to capture the differences between poor- and good-quality teaching. On the other hand, the independent variables are heterogeneous. Teacher degrees may reflect quite rigorous preparation or may be available from a “diploma mill.” The same is true for teacher in-service training and professional development, he added; they vary so much that the benefits of excellent programs might be obscured in data that also include poor-quality programs.
At the same time, teachers and students are not independent—children are not randomly assigned to preschool classrooms and teachers are not randomly assigned to programs, and these selection biases significantly complicate studies of the contributions teachers make to children’s outcomes. Teachers who work together also influence one another. A study that focused on the lead teachers without taking into account the influence of coteachers would likely miss the effects of their interactions. Moreover, the varying ways in which centers’ or states’ policies regarding staffing are implemented may also bias results.
Another set of issues relates to the factors that influence how much teacher education is typical in a program. The curriculum, the director’s views, the availability of services and supports (e.g., child study teams or coaching), and the program’s educational goals are likely to affect the program’s attractiveness to teachers with different credentials and its hiring objectives. Compensation and working conditions are also important, Barnett stressed. Studies that hold those constant when examining the effects of teacher qualifications, he said, “make no sense.”
Barnett suggested that research on “packages,” that is, programs with particular sets of characteristics, or policy regimes, might be more useful because the features are so interactive. Certainly program effects do vary substantially, as the data in Table 4-1 show. However, data from New Jersey’s Abbott Prekindergarten program show how a change in policy can
TABLE 4-1 Outcomes for Three Preschool Programs
|8 State Pre-Ks|
|Avg||Lowest||Highest||Perry Preschool||Tulsa Pre-K||Head Start (adj)|
|Print awareness||.97||.56||1.49||NA||.99||.25 (.34)|
affect several child outcome measures (Frede et al., 2009). In 1999, the program raised its requirements for teachers, implemented a research-based curriculum, added master teachers, restricted class size to 15, incorporated high standards and accountability measures, and adopted a continuous improvement system. Measures of classroom quality, including the quality of the language, literacy, and mathematics environments, showed substantial improvements between the 2002–2003 and the 2008–2009 school years. The goal was overall improvement; improvement in teacher qualifications was just one part of the plan.
“We know many preschool programs are not delivering the desired results,” Barnett concluded. “Inadequately prepared teachers are one likely cause, but it is not by itself the thing that’s going to make a difference.” Particular qualifications might bring teachers who can be highly effective, he observed, but those teachers will not necessarily be highly effective unless other things are in place. Teacher pay and working conditions, supports for teachers, the curriculum, and state and local policies have a material effect on the way teachers function in their jobs. In his view, teachers are best able to be effective when they are an integral part of a process of continual improvement. Overall, Barnett summarized his policy recommendation regarding whether teachers should have B.A. degrees in the following way: “To me when we’re dealing with policy, when we’re unsure, I think we have to ask ‘what is the cautious thing to do?’ I think moving away from programs with highly educated, well-paid teachers is an incautious thing to do given the evidence.”
Workshop participants added many comments. One observed that current standards are very low, both for the individuals who enter the profession and for the level and nature of the skills and behaviors those
workers are expected to have. Others highlighted the importance of effective management and well-trained professional development providers, and the need for both knowledge-oriented and practice-focused experiences in higher education teacher preparation programs. Others pointed to the need for well-defined goals, clearly agreed-on outcome measures, and timely, clear guidance about how to proceed.
Several emphasized the point that Barnett stressed—that no one single element or strategy will produce the desired quality or outcomes. However, others cautioned that evaluating combinations of factors will limit the ability to determine which of them are essential to producing the desired outcomes. One participant suggested that observational studies in addition to randomized controlled trials could be useful. Another noted that randomized experimental research focused on teacher characteristics and preparation was emerging, such as work being conducted by Susan Neuman at the University of Michigan.
Another discussion focused on the value of data-driven decision making as it relates to teacher qualifications. Although this is an important goal, several participants noted that measures of outcomes for children are very limited. “We need clearly agreed-upon outcomes that go way beyond the Peabody Picture Vocabulary Test and some of the pathetically narrow measures that we’ve been using in order to have the kind of robust research designs that can give us some confidence that we’re on the right track,” one observed. Unfortunately, applying a K–12 model for measuring outcomes is not likely to serve the early childhood context well, others noted. Social and emotional development, for example, is difficult to measure but is “an extremely important part of what we hope to stimulate with whatever kinds of care experiences children have,” another noted. Furthermore, the kinds of skills, activities, and goals that are in play for infants, toddlers, and preschoolers are different, and those distinctions are not clear in the available measures.
Diversity and the Early Childhood Care and Education Workforce
Other teacher and caregiver characteristics may also influence outcomes for children, explained Ellen Frede, codirector of the National Institute for Early Education Research. She discussed the available research pertaining to two questions: (1) Is there a demographic match between caregivers5 and the children in their care; and (2) Does a demographic match make a difference in terms of outcomes for children? The first question is not easy to answer, Frede explained. The available data pro-
5 This section of the report uses the term “caregivers” as it was used in the workshop presentation.
TABLE 4-2 Demographic Matches Between Caregivers and Children in New Jersey, U.S. Census Data
|2010 Census||0–5 years||All Ages||Preschool K Teachers|
|White alone||74.5||79.6||~ 80|
|American Indian, Alaskan Native||1.4||1|
|Native Hawaiian/Pacific Islander||0.3||0.2|
SOURCE: Frede, 2011.
vide relatively little detail about caregivers’ characteristics, and even less information that could be used to determine the matches between caregivers and children. Consequently, the second question is also difficult to answer, but because of its potential importance for policy it should not be overlooked, in Frede’s view.
U.S. Census data provide a big-picture look, as shown in Table 4-2. Frede noted first that these data suggest it would be surprising to find a close match between caregivers and children, given that the proportion of children ages 0 to 5 who are minorities is greater than the minorities’ proportion of adults in the general population.6 The Head Start program has national enrollment data that also illuminate the issue, as shown in Table 4-3. That program has achieved a higher-than-average match rate, although the adults are more predominantly white and the children are more predominantly minority. Frede also described data for family child care in New Jersey, which shows a pattern similar to that in the Head Start data. A few studies from the 1990s looked explicitly at the matches between children and caregivers, as shown in Table 4-4, and, like the others, found lower matches for minority children.
The most significant discrepancy, she pointed out, is between His-
6 Frede cautioned both that the race/ethnicity categories and the classifications for the child care and education workforce are those defined by the Census and may obscure aspects of the demographic matches.
TABLE 4-3 Demographic Matches Between Children and Caregivers, Head Start 2009–2010
|2009–2010 Head Start Program Information Report|
|Enrollment Statistics Report–National Level|
|Biracial or multi-racial||8%||4%|
|American Indian, Alaskan Native||4%||4%|
|Native Hawaiian/Pacific Islander||1%||1%|
|Hispanic or Latino origin||34%||25%|
SOURCE: Frede, 2011.
TABLE 4-4 Demographic Match Between Teachers and Children, Research Findings
|Data Source||Euro-American||African American||Hispanic/Latino|
|NICHD (Burchinal and Cryer, 2003)||86%||70%||34%|
|CQ&O (Burchinal and Cryer, 2003)||77%||47%||42%|
|Multi-State Home-School Match (Barbarin et al., 2010)||90%||32%||8%|
NOTE: CQ&O: Cost, Quality, and Child Outcomes in Child Care Centers Study; NICHD: National Institute of Child Health and Human Development.
SOURCE: Frede, 2011.
panic or Latino children and teachers. Demographics may make a difference if language differences impede communication and effective education. She presented some additional New Jersey data showing that providers of state-funded prekindergarten have larger percentages of bilingual teachers than do public school prekindergarten classrooms, a point discussed further below.
Why might the ethnic and racial matches between teachers and children matter? Research and theory have both suggested possible reasons, Frede explained. An older theory, based in qualitative research, was that discontinuity between home and school could be confusing to children (Silvern, 1988). It might affect teachers’ attitudes toward children and also
affect children’s identity development. For example, children who have never had teachers of their own ethnic background might come to believe intuitively that the group from which teachers came was supposed to have more power.
Researchers have also suggested that differences in discourse patterns among ethnic groups may be important in this context. For example, Heath (1989) found that Euro-American teachers may use discourse patterns that are characterized by questions for which the adult knows the answer, while African-American parents were more likely to ask their children questions to which they actually wanted an answer. Such differences may yield misunderstanding between teachers and children.
Burchinal and Cryer (2003) have also explored the relationship between demographic matching and outcomes for children. They found that, regardless of race or ethnicity, what mattered most for outcomes were caregivers’ sensitivity and the amount of stimulation they provided. These two factors predict school success across race and ethnicity. Neither a race/ethnicity match between child and caregiver nor a match of race/ ethnicity and child-rearing beliefs between caregiver and parents had strong influences on outcomes for children, Frede explained. However, the researchers also stressed that there were many ways to provide sensitive and stimulating care while avoiding a disconnect with families.
Another study (Barbarin et al., 2010) also examined the effects of matches, using a slightly different description of teacher characteristics. They found, in a public prekindergarten setting, that fostering independence but providing high levels of support promoted kindergarten readiness. Even when families hold different beliefs about child rearing, teachers who follow that model improve children’s readiness, though to a slightly lesser degree. In other words, Frede explained, outcomes are best if both teachers and parents use this approach, but if either one does, children will benefit.
Several studies suggest that mismatches might be a more significant issue for children who are English-language learners, Frede pointed out (Barnett et al., 2007; Durán et al., 2010; Farver et al., 2009; Gormley, 2007; Kersten et al., 2009). These studies collectively demonstrate that, in general, attending a high-quality preschool improves outcomes for English-language learners. The researchers also found that when the preschool environment uses both English and the child’s home language, the children’s proficiency in the home language improves, and they do not lose their progress in English.
Having a bilingual teacher is clearly a prerequisite for dual-language instruction, Frede added, but Freedson (2010) found that most dual-language children do not have a teacher who speaks their language or has specialized knowledge of this type of instruction. Moreover, Frede
reported, teacher preparation programs neither routinely offer substantive coursework in linguistic or cultural diversity nor require students to learn a second language. Frede also found that bilingual teachers tend to use their (and children’s) home language primarily for commands and discipline.
Frede drew several conclusions from these findings:
The match between children and teachers in early childhood education settings with respect to race, ethnicity, and language is largely unknown but is “clearly not one-to-one”;
Perfect matching is likely not possible and, if it were, would result in de facto linguistic or racial segregation;
What research exists suggests that demographic matches between teachers and children are not necessary for effective care and education, and that sensitive and supportive teachers are more predictive of beneficial outcomes; and
To optimize learning for dual-language learners and increase bilingualism, more bilingual teachers need to be trained to support dual-language learning.
Frede noted that better data will be needed to shed more light on this issue. In the meantime, she believes that a few steps would make high-quality, dual-language education more widely available to children under age 5. States should require teacher preparation programs to improve bilingual education; in fact, this should be a criterion for receiving state and federal funds, she argued. She also believes that teachers should have access to professional development that fosters sensitive and stimulating interactions with children and an approach that promotes independence while providing support.
The characteristics teachers and caregivers bring to the job and develop in the course of their careers are very important. The working conditions for those employed in ECCE, however, also influence the decisions prospective workers make about entering and remaining in the workforce, as well as the quality of their performance. Deborah Phillips, professor of psychology at Georgetown University, and Marcy Whitebook, senior researcher and director of the Center for the Study of Child Care Employment at the University of California, Berkeley, described research on the environmental factors that influence teachers’ experiences and outcomes for children.
The Role of Stress for Children and Caregivers7
Phillips drew from research on the neurobiology of stress, individual differences in temperament, and child care itself to develop a portrait of the intersecting factors that affect the workforce. She briefly reviewed the evolution of research and thinking about these influences. The first wave of research on child care, she explained, focused on comparisons between children cared for at home by their mothers and those in child care. But researchers quickly realized that child care arrangements vary widely and began looking for the ingredients that make it effective.
The next phase of research explored ecological models. Researchers attempted to control for the impacts of home environments and parenting, to control for selection bias in their study samples (i.e., unobserved factors that may influence the type of children served by different child care settings), and to focus not only on the setting for child development, but also on the setting for adult workers. More recent work examines interactions between child care and parenting and the relationships between parents and caregivers. Current work is also focusing on the theory that quality in child care is not necessarily linear, and that certain thresholds may need to be reached before the positive benefits of child care can be demonstrated.
At the same time, advances in research on child development—and especially early brain development—and on the neurobiology of stress have shed light on the critical role played by caregivers. The developing brain, Phillips explained, is designed to “greedily recruit from the environments and experiences that surround it as they shape its emerging architecture and neurochemistry.” Early experiences “literally mold the brain,” but it will recruit both supportive and damaging experiences. A critical element in this process is the development of the structure that manages the brain’s and body’s response to stress (the hypothalamic-pituitary-adrenal [HPA] axis). This system produces cortisol, a steroid hormone that plays a critical role in adaptation to stressful stimuli. Researchers have found that regular and sustained activation of this system in animals early in life is associated with increased fearfulness and impairment of the capacity to regulate responses (Gunnar, 2008; NSCDC, 2005). In particular, these animals’ responses to novelty, struggles for dominance, and reactions to social threats are affected.
Efforts to understand the circumstances under which activation of the HPA system has adaptive or non-adaptive effects have led to the development of a three-part classification of stress, Phillips explained: positive stress, tolerable stress, and toxic stress (NSCDC, 2005). Positive stress is
7 This section of the report uses the term “caregivers” as it was used in the workshop presentation.
mild and temporary, and promotes the developing capacity to respond to the inevitable stress of life. Tolerable stress is more enduring and intense, but does not compromise the developing HPA system, often because of the buffering presence of a dependable and responsive adult. Toxic stress includes stress in the absence of protective factors (e.g., a supportive adult) or severe abuse, neglect, or prolonged maternal depression that deprives the child of the secure presence of a loving adult.
Although every developing child would be affected by toxic stress, Phillips explained, some are more vulnerable than others. Researchers believe that children’s innate temperaments can affect their capacity to regulate their stress response systems—and some are characterized by “negative reactivity” (Martin and Fox, 2008; Rothbart et al., 2006). Different temperaments can be detected in babies as young as age 4 months. When they are presented with intense stimuli, such as loud noises, busy pictures, and multiple people, some “freak out—they arch their backs and cry”; they are clearly miserable. Babies at the other end of the spectrum may thrive on the same level of stimulation. As they grow, a share of the infants with highly negative reactions remains easily stressed and exhibits high anxiety and high cortisol levels in stressful social situations.
These children, Phillips explained, are particularly sensitive to their contexts, including variations in the quality of the care they experience in their first months and years. When these children do not receive dependable, sensitive care, they tend to display one of two behavioral profiles: fearful, anxious retreat from stressful situations, or aggression and lack of behavioral control. When they do receive dependable, sensitive care, however, they are usually able to develop well-calibrated stress response systems.
This work, Phillips noted, has focused researchers’ thinking about the conditions under which child care may compromise development, and the conditions under which children thrive in child care. This perspective, she added, “really ups the ante on how we think about the adults who are caring for young children. What responsibilities are we really placing in their hands? It goes way beyond … dressing, feeding, bathing, and overseeing play.” Evidence has long been clear that child care varies widely in quality and that its quality matters a lot to children’s outcomes, she noted. Now, new data indicate that poor-quality child care appears to be a highly stressful environment for a particular group of children. They display rising levels of cortisol over the course of the day in child care that they do not exhibit at home—whereas the normal pattern is for cortisol levels to be highest at the start of the day and to drop in the course of the day.
Emerging evidence indicates that this circumstance may profoundly affect these children’s development. Inhibited, fearful, socially reticent children seem to be the most vulnerable, especially at ages 2 and 3 (Phillips
et al., 2011). “These are children who are on the periphery of the action a lot of the time,” she explained. “They are seen watching rather than participating in play and social activities, but they’re not watching happily. They are feeling left out. They are really anxious. They’re self-soothing. They’re clinging to adults. It’s really painful to observe these children.” Studies are beginning to show that poor-quality care is associated with a higher incidence of stress responses in children, whether in center- or home-based programs (Phillips et al., 2011). The elevated cortisol has also been associated with increased frequency of infections and lower antibody levels.
What is still missing is a clear way to determine the threshold at which child care moves from stressful to beneficial. Intrusive, overcontrolling care, she added, appears to have a negative influence, while warm, sensitive care has a positive effect. Emerging research suggests specifically that a secure attachment between the child and his or her lead teacher seems to protect against the rising cortisol levels (Badanes et al., 2011; Gunnar et al., 2011). Existing measures of child care quality are not bad, Phillips observed, but they are “missing something really important” because they do not capture this aspect of care. Another area of concern is the mental and emotional state of the caregivers and teachers themselves, Phillips explained. Some research has shown that 16 percent of family child care providers and 22 percent of center teachers have scores above the threshold for genuine depression (Whitebook et al., 2004). In both settings, those who care for low-income children have the highest rates of depression. Some have reported links between caregiver depression and intrusive, overcontrolling care that has a negative effect. Those who care for children on their own are most vulnerable to depression that can interfere with the quality of care. The emerging evidence suggests that, like children, some caregivers exhibit rising cortisol levels in the course of the day, and this effect is negatively associated with child well-being.
This body of work, in Phillips’ view, makes it clear that caregivers need to know how to help children navigate and manage peer interaction and how to help them develop the self-regulatory and executive functioning skills that underlie both social competence and cognitive and language development. Caregivers need to be perceptive about and sensitive to individual differences among children. Caregivers with these characteristics and skills are particularly important for children with any kind of vulnerability related to temperament or special needs or poverty. Phillips also emphasized that child care for all children must be understood as an intervention—one that will affect their developmental trajectories, “either in an adaptive or a compromised direction.”
Stability and Stress
Whitebook presented data on turnover and other aspects of the workplace that can contribute to teacher and caregiver well-being and quality. Child care environments vary tremendously, she observed. Worker turnover, stress, and skill levels are intertwined, and these interactions are important to center quality (she noted that most of the research in this area focuses on centers rather than home-based care).
Job turnover can be positive, Whitebook noted, when it means that staff who are not very effective are leaving the field, or that individuals are leaving jobs that do not reward their investment in education and professional development. Conversely, stability is not a positive sign when workers stay because they believe they have no other opportunities, receive no encouragement to improve, or do not experience consequences for failure to improve. In any case, job turnover in child care is very high, as the right-hand column in Table 4-5 shows. Some evidence suggests it is highest for those who work with infants and toddlers, who are also the most vulnerable to the consequences of turnover. For example, a study of child care workers in California showed that turnover was 29 percent in centers that serve infants and toddlers as well as older children, and 20 percent for those that serve only older children (Whitebook et al., 2006). Turnover was highest in centers that serve children in subsidized programs. Turnover in family child care has been less well researched, but one Illinois study found a rate of 25 percent turnover over a 15-month period for licensed family providers (Fowler et al., 2008). Another California study found that among providers who were receiving a subsidy to care for children whose mothers had just stopped receiving welfare benefits, 43 percent left in the course of a year (Whitebook et al., 2003).
Numerous factors may influence turnover in any occupation, Whitebook explained. Poor compensation, poor hiring decisions, and poor working conditions are the most frequent causes. In the early childhood context, wages appear to be an especially significant factor. Table
TABLE 4-5 Wages and Stability
|Occupation||Mean Hourly Wage||Turnover Rate|
|Home health aides/nurse’s aides||$10.39||18%|
|Child care workers||$10.07||29%|
|Food counter workers||$ 9.13||42%|
SOURCE: Whitebook, 2011; Based on U.S. Department of Labor, Bureau of Labor Statistics, 2009.
4-5 also shows the hourly wages for child care workers and several other occupations, and demonstrates an inverse relationship between salary and turnover rate. Despite the fact that unemployment is currently very high and states and other jurisdictions are struggling with serious budget shortfalls, Whitebook stressed, relative compensation for child care workers is still important. The Department of Defense, she noted, has demonstrated that it was possible to sharply reduce turnover in the military child care system without passing on the labor costs to parents. Faced with a turnover rate of 48 percent in 1989, the Department of Defense instituted a new pay and training plan. They made pay for child care jobs commensurate with that for other jobs with similar requirements for education and training. The turnover rate was reduced to 24 percent by 1993 (Campbell et al., 2000).
Children experience turnover in their teachers and caregivers as loss, Whitebook explained, and it affects the quality of their care. A 1989 study (Whitebook et al., 1990) showed, for example, that children in centers with low quality and high turnover have poorer language and social development. Other work (Whitebook et al., 2001, 2004) showed that centers that were able to improve their ECERS scores, gain accreditation, and sustain those improvements were also those that had the lowest turnover and paid the highest wages. She added that directors and teachers “talk about turnovers as the time sponge, the energy drain, or the plague, depending on how bad it is.” Finding a replacement for the worker who has left may take 3 to 6 weeks. Children may need to be regrouped, and “conversation with children stops because everybody is just trying to keep their heads above water.” Because of pressure to fill the position quickly and meet ratio requirements, directors may end up hiring people who are less skilled (the “warm body” syndrome), and quality declines. Often the result is additional turnover, as well as stress for teachers, caregivers, parents, and directors. Turnover among directors is also a problem, which can both exacerbate and be exacerbated by staff turnover.
Among caregivers who leave child care jobs, Whitebook noted, data show that about 42 percent move to other centers, 8 percent move to resource and referral agencies, 7 percent move to K–12 education, and 21 percent leave the field altogether (Whitebook et al., 2001). A survey of people working in resource and referral agencies found that about 50 percent had a background working in child care—and that two-thirds of those people said they have left direct child care because of the pay (Whitebook et al., 2010).
Even though having any job in a time of high unemployment is better than none, working for very low wages is nevertheless stressful. One study (Whitebook et al., 2011) has shown that only 44 percent of child
care workers would be able to support themselves and their families on their own wages alone. Only about a third of child care workers have employer-paid benefits or paid sick leave. These conditions are stressful, and, as Phillips discussed, stress and depression among caregivers can have negative effects on children. Whitebook added that working with stressed and depressed colleagues can be stressful in itself and undermine the professional culture in a center.
In general, she observed, the turnover problem is an aspect of the broader challenge of ensuring that child care and education programs are staffed by people who have the appropriate skills (e.g., dual-language training, special education experience), the credentials, and the social support necessary. The work environment is a critical ingredient in quality child care, she believes, but most states use quality rating and improvement systems that do not really address what is necessary to create a positive work environment. A few states (Colorado, Illinois, and New Mexico) have begun to focus on workplace quality, but most have not yet done so, Whitebook and colleagues found (Austin et al., 2011). Although further research will be needed to support firm conclusions, Whitebook and her colleagues identified several working conditions that are likely to be important:
A workplace that is a learning environment in which workers have the opportunity to reflect on and discuss their work;
A workplace in which people are empowered to make changes as they learn new strategies and skills;
The resources necessary for adult well-being, such as mental health referrals for workers who are having problems or paid leave for illness; and
Adequate financial rewards.
The well-being of the adults in early childhood settings—their living and working conditions—are not a distraction from the needs of children, Whitebook added, but rather the cornerstone of children’s environment, and thus an essential determinant of how well children are going to do.
Participants identified additional workforce issues that also merit further investigation. One described a “disconnect” between (1) what researchers have learned about children’s development and the care they need, and (2) the way the field conceptualizes the skills the workforce needs and approaches recruitment. The low status the occupation cur-
rently has, together with the low wages, is likely to continue to limit its appeal to academically successful college graduates. Others worried that as expectations for child care and education are raised, the workforce who can deliver that level of competence is just not available. The workforce needs a range of financial as well as non-monetary supports if quality is to improve, they suggested.