## Abstract

Historically, the early childhood care and education (ECCE) workforce has been characterized as a low-education, low-compensation, low-stability workforce. In recent years, considerable investments have been made to correct this, but we lack evidence about the extent to which these investments were accompanied by changes in the characteristics of the workforce. Using nationally representative data, we find that the historical characterization of the ECCE workforce continues to apply. However, we also find that the average educational attainment, compensation, and stability of ECCE workers increased substantially from 1990 to 2010. Surprisingly, the shift in the composition of the ECCE workforce toward more regulated settings and away from home-based settings is not the primary driver of these changes. Contrary to our expectations, gains within the home-based workforce are the primary drivers, though the education and wages of home-based workers remain substantially lower than among formal-care workers.

## Introduction

In the United States, most children under age 5 years receive regular care by someone other than their parents (Bassok 2010; U.S. Census Bureau 2010). Early childhood experiences play a central role in shaping subsequent developmental trajectories, and the impact of these early experiences depends largely on the quality of caregivers and teachers (Shonkoff and Phillips 2000; Peisner-Feinberg et al. 2001; Knudsen et al. 2006; Hamre and Pianta 2006; Fox, Levitt, and Nelson 2010).

Growing recognition of the importance of early childhood care and education (ECCE) in general, and of ECCE providers in particular, has heightened policy interest in strengthening the quality of the ECCE workforce. In 2011, the federal government funded the Race to the Top Early Learning Challenge, a competitive grant program to support states’ efforts to improve early childhood education programs, and identified “supporting a great early childhood education workforce” as one of five key areas of reform (USDOE 2011, p. 6). The latest reauthorization of the federal Head Start program required that 50 percent of Head Start teachers hold a bachelor's degree (BA) in child development or a related field by 2013 (Barnett et al. 2010). Further, twenty-five states are operating or developing Quality Rating and Improvement Systems (QRIS) to assess and improve the quality of ECCE, and many of these QRIS programs offer financial incentives to providers that invest in their employees’ education and training (Tout et al. 2010).

Despite the interest in the improvement of this sector, we know relatively little about the current state of the ECCE workforce, and even less about the extent to which this workforce has changed over time. It is well documented that the ECCE workforce is characterized by low levels of education, wages, and stability (IOM and NRC 2012; Howes, Phillips, and Whitebook 1992; Helburn 1995; NICHD Early Child Care Research Network 2000; Vandell and Wolfe 2000). For instance, the average annual income of paid ECCE workers in 2009 ranged from $11,500 for those working in a child's home to$18,000 for preschool teachers (GAO 2012).1 However, although studies have documented the low education, wages, and stability of ECCE workers using a variety of data sources, the diverse and dispersed nature of the industry makes systematic analysis difficult. A recent report of the National Research Council describes how the lack of comprehensive data tracking the characteristics of the ECCE workforce seriously limits policy makers’ efforts to facilitate change or track improvements over time (IOM and NRC 2012).

Over the past 20 years utilization of “formal” ECCE services such as preschool, state pre-kindergarten, and Head Start has increased rapidly. This increase has been accompanied by a decline in the share of workers employed in more “informal” home-based settings, such as family childcare homes (Bassok, Fitzpatrick, and Loeb 2012). Given that the home-based sector faces much less stringent regulations than the formal sector, and is often singled out for providing the lowest-quality care, the shift toward formal care may have translated into overall improvements in the ECCE workforce over time (Bassok et al. 2012). Unfortunately, attempts to describe the evolution of the ECCE workforce have been limited due to the paucity of data that allows for reliable comparisons of the workforce over time (Saluja, Early, and Clifford 2002; Brandon and Martinez-Beck 2006; Kagan, Kauerz, and Tarrant 2007).

The few studies that have examined the evolution of the ECCE workforce over time actually suggest that the qualifications of the workforce have either changed only modestly or have declined (Whitebook et al. 2001; Saluja, Early, and Clifford 2002; Herzenberg, Price, and Bradley 2005; Bellm and Whitebook 2006). These studies, however, do not use nationally representative data and/or focus only on a single sector of the ECCE industry, typically childcare centers. The lack of knowledge about changes within the home-based workforce represents a particularly relevant gap in the literature, given that this sector accounts for about a third of the national ECCE workforce (GAO 2012).

In this research brief we make use of nationally representative data that encompass workers in all three ECCE sectors—centers, homes, and schools—to address three questions:

1. What are the characteristics of the ECCE workforce as of 2010?

2. How did the characteristics of this workforce change between 1990 and 2010?

3. To what extent are the overall changes driven by a change in the relative importance of each sector (centers, homes, schools), and to what extent are they explained by changes in the characteristics of the workforces within each sector?

We focus on four outcomes to gauge the well-being of the ECCE workforce and plausibly proxy for ECCE quality: (1) the educational attainment of workers; (2) their compensation; (3) the extent to which workers exit the industry over a year; and (4) the occupational prestige of those who enter the ECCE workforce each year from other occupations. Improvements along these dimensions are likely to reflect an increased ability to attract and retain qualified workers into the ECCE industry, and in turn may imply higher quality experiences for young children.2

We find that the low-education, low-compensation, high-turnover characterization of the national ECCE workforce continues to be valid. At the same time, we show that the average educational attainment and compensation of the ECCE workforce increased between 1990 and 2010, and that turnover from the ECCE industry decreased substantially. Our results differ from earlier studies that highlight negative or stagnant trends in the ECCE workforce. These differences are likely explained by our focus on a more recent period of analysis and our use of national data, including workers from all three ECCE sectors. We also show that changes in the characteristics of the national workforce are mostly explained by changes in the characteristics of workers within each sector and less so by the shift toward center- and school-based settings. Surprisingly, we find that changes along all dimensions analyzed were most pronounced among home-based workers.

## Data

We analyze data from the March Supplement of the Current Population Survey (CPS), a nationally representative household survey that is administered every month by the U.S. Census and the Bureau of Labor Statistics. Using the Census 1990 and 2002 Industry and Occupational Codes, we identify ECCE workers and disaggregate this broad group into center-, home-, and school-based workers. We purposefully implement a broad and inclusive definition of the industry. Specifically, our center-based category includes all workers who (1) are not self-employed; (2) work in either the “child day care services” industry, or have childcare occupations (e.g., “childcare workers,” “pre-kindergarten or kindergarten teachers,” “early childhood teacher's assistants”); and (3) work in an industry other than “elementary and secondary schools,” “private households,” “individual and family services,” or “family childcare homes.”3 Our definition of the home-based ECCE workforce includes (1) all self-employed individuals who report that they work in the “child day care services” industry; (2) all those employed in the “family childcare homes” industry; (3) those who have childcare occupations (e.g., “childcare workers,” “private household childcare workers,” “pre-kindergarten or kindergarten teachers,” “early childhood teacher's assistants”) and are employed in the “private households” or “individual and family services” industries; and (4) those who have childcare occupations and are self-employed in other industries except for “elementary and secondary schools.”4 Finally, we define the school-based ECCE workforce as “pre-kindergarten and kindergarten teachers” and “early childhood teacher assistants” employed in the “elementary and secondary schools” industry. We observe whether each respondent was an ECCE worker in the week of reference and whether their longest job in the previous calendar year was an ECCE job.

The workforce characteristics that we analyze are measured as follows:

1. Educational attainment: The CPS collects information about each household member's highest level of education as of the week of reference. In keeping with prior studies, we describe changes in the share of ECCE workers with less than a high school degree, exactly a high school degree, at least some college education but no BA, and at least a BA.5

2. Compensation: We observe each individual's annual earnings from the longest job held in the previous calendar year. We describe the mean annual earnings of those whose main job in the previous calendar year was an ECCE job. We also estimate the hourly earnings of these workers, but here restrict our analysis to those who were full-year workers in the previous calendar year.6 We express both earnings variables in 2010 dollars.

Individuals also report whether any employer helped pay for a pension and/or health plan in the previous calendar year. We use this information to construct the share of ECCE workers who received this non-salary form of compensation. Here we restrict our sample to workers whose main job in the previous calendar year was an ECCE job and, in order to be sure the benefits were received from an ECCE employer, we include only those workers who reported they had only one employer in the previous calendar year.7

1. Year-to-year industry turnover: To measure ECCE industry turnover rates, we exploit the fact that the CPS provides information about an individual's industry and occupation both in the week of reference and for the longest job held in the previous calendar year. Among individuals whose main job in the previous calendar year was an ECCE job, we estimate the industry turnover rate as the share of those who were no longer in the ECCE workforce during the week of reference. An analogous method is used by Harris and Adams (2007) to measure turnover from elementary and secondary teaching. We can calculate industry turnover with the CPS from 1990 to 2010. Our measure only captures whether individuals remained in the ECCE workforce; among those that remain, we cannot distinguish whether individuals changed jobs. Thus, year-to-year industry turnover is a lower bound estimate of the level of instability experienced by children.

2. Occupational prestige of entrants into the ECCE workforce: We combine the information on a worker's occupation provided by the CPS with a commonly used methodology developed by Charles Nam and colleagues (Nam and Boyd 2004; Nam 2000), to create a variable that assigns each new entrant to the ECCE workforce a score based on the occupational prestige of their previous job. A particular occupation's prestige score is constructed by comparing the median earnings and educational attainment of workers in that occupation vis-à-vis the earnings and education of workers in all other occupations. An occupation's score can range from 0 to 100, and reflects the percentage of individuals in the labor force who are in occupations with combined levels of education and earnings below that occupation. We use these scores to examine the average occupational status of individuals whose main job in the calendar year before the survey was outside the ECCE industry, but who were ECCE workers in the week of reference. Increases in this occupational measure over time imply that those who are entering the ECCE workforce are coming from better educated and better paid occupations than those who were entering the workforce in previous years.

As researchers have long pointed out, existing data sets fail to fully and accurately capture the complexity of the ECCE workforce over time (IOM and NRC 2012; Bellm and Whitebook 2006; Phillips and Whitebook 1986). Although the CPS is well suited for nationally representative analyses tracking trends over time, it has a number of key limitations: (1) it relies on self-reported data on employment, and therefore likely excludes many unpaid ECCE workers and some paid family, friends, and neighbors who take care of children but do not report child care as their occupation; (2) it does not enable us to distinguish between preschool and kindergarten teachers, or, more generally, to distinguish ECCE workers by the age of the children they serve; and (3) it does not collect detailed data that are relevant to characterize ECCE workers, such as the level of ECCE-specific training, the responsibilities they have, or the quality of their interaction with children. We return to these limitations in discussing the generalizability of our results.

## Methods

To address our first and second research questions, we present the variables of interest in 2010, and discuss their change over the period 1990–2010. We assess whether trends in the ECCE workforce differ from broader trends in the economy by comparing changes in that workforce to changes among two comparison groups: all female workers and low-wage workers.8 To address the third research question, two sets of simulations allow us to disentangle the extent to which the overall changes in the ECCE workforce are explained by an increase in the relative size of the more regulated ECCE sectors or by changes in the workforce within each sector.9 Given the relatively small sample size of the CPS in each year, for all analyses we use three-year moving averages to increase the precision of our estimates.

## Results

### The ECCE Workforce as of 2010

We find that the “low-education, low-compensation, high-turnover” label continues to be a valid characterization of the 2.2 million ECCE workers represented in our sample. As shown in table 1, in 2010, nearly 40 percent of the ECCE workforce had at most a high school degree and a third of the workforce had some college but no bachelor's degree. In 2009, the average ECCE worker earned an annual income of $16,215 and an hourly wage of$11.70, and only 28 percent of ECCE workers received a pension and/or health benefits from their employer.10 Worryingly, about a fourth of those workers who had been employed in the ECCE industry in 2009 were no longer in that industry by 2010. Further, our analysis of the occupational prestige of entrants suggests that ECCE was a relatively unattractive industry to enter in 2010, attracting individuals from occupations that on average had lower levels of education and earnings than three fifths of the country's labor force.

Table 1.
Evolution of the ECCE Workforce, and Comparison to Female and Low-Wage Workers (1990–2010)
199220102010 vs. 1992
Distribution of the workforce by educational attainment
ECCE workers
Less than high school 21.4 11.5 −9.9*
High school degree 31.5 26.9 −4.6*
Some college or associate's degree 26.1 33.3 7.2*
At least a bachelor's degree 20.9 28.4 7.5*
Female workers
Less than high school 11.5 8.1 −3.4*
High school degree 36.0 26.4 −9.6*
Some college or associate's degree 29.2 31.9 2.7*
At least a bachelor's degree 23.2 33.6 10.4*
Low-wage workers
Less than high school 20.5 17.0 −3.5*
High school degree 38.9 33.5 −5.4*
Some college or associate's degree 26.7 31.1 4.4*
At least a bachelor's degree 13.9 18.4 4.5*
1990 2009 2009 vs. 1990
Mean annual earnings of all workers (at 2010 dollars)
ECCE workers 10,746 16,215 51%*
Female workers 24,427 30,629 25%*
Low-wage workers 18,266 21,298 17%*
Mean hourly earnings of full-year workers (at 2010 dollars)
ECCE workers 8.8 11.7 33%*
Female workers 16.3 19.0 17%*
Low-wage workers 13.4 14.2 6%*
Share of workers with pension and/or health benefits paid at least partly by the employer
ECCE workers 19.0 28.0 9.0*
Female workers 56.4 57.9 1.5*
Low-wage workers 42.5 42.2 −0.3
1990 2010 2010 vs. 1990
Industry turnover rate
ECCE workers 32.9 23.6 −9.3*
Female workers 24.7 17.9 −6.8*
Low-wage workers 26.5 19.1 −7.4*
Average occupational prestige in the year before entering the workforce
ECCE workforce enterers 37.6 42.3 4.7
Low-wage workforce enterers 41.8 42.0 0.2
199220102010 vs. 1992
Distribution of the workforce by educational attainment
ECCE workers
Less than high school 21.4 11.5 −9.9*
High school degree 31.5 26.9 −4.6*
Some college or associate's degree 26.1 33.3 7.2*
At least a bachelor's degree 20.9 28.4 7.5*
Female workers
Less than high school 11.5 8.1 −3.4*
High school degree 36.0 26.4 −9.6*
Some college or associate's degree 29.2 31.9 2.7*
At least a bachelor's degree 23.2 33.6 10.4*
Low-wage workers
Less than high school 20.5 17.0 −3.5*
High school degree 38.9 33.5 −5.4*
Some college or associate's degree 26.7 31.1 4.4*
At least a bachelor's degree 13.9 18.4 4.5*
1990 2009 2009 vs. 1990
Mean annual earnings of all workers (at 2010 dollars)
ECCE workers 10,746 16,215 51%*
Female workers 24,427 30,629 25%*
Low-wage workers 18,266 21,298 17%*
Mean hourly earnings of full-year workers (at 2010 dollars)
ECCE workers 8.8 11.7 33%*
Female workers 16.3 19.0 17%*
Low-wage workers 13.4 14.2 6%*
Share of workers with pension and/or health benefits paid at least partly by the employer
ECCE workers 19.0 28.0 9.0*
Female workers 56.4 57.9 1.5*
Low-wage workers 42.5 42.2 −0.3
1990 2010 2010 vs. 1990
Industry turnover rate
ECCE workers 32.9 23.6 −9.3*
Female workers 24.7 17.9 −6.8*
Low-wage workers 26.5 19.1 −7.4*
Average occupational prestige in the year before entering the workforce
ECCE workforce enterers 37.6 42.3 4.7
Low-wage workforce enterers 41.8 42.0 0.2

Notes: *Denotes change with respect to 1990 or 1992 is statistically significantly different from zero at the 5% level.

Changes in the share of workers by educational attainment, the share with pension and/or health benefits, and the industry turnover rate are measured in percentage points; changes in annual and hourly earnings, as a percent change; and changes in the average occupational prestige score of those entering the ECCE workforce, in percentiles.

Source: Authors based on the March Supplement of the Current Population Survey.

The disaggregated results shown in table 2 highlight stark differences across sectors. In 2010, about 56 percent of ECCE workers were employed in center-based settings, 26 percent in home-based settings, and 18 percent in schools. Consistent with evidence from prior studies, we find that the school-based workforce exhibits the highest levels of formal education, compensation, and stability, and the home-based workforce exhibits the lowest. The center-based workforce falls in the middle but is more similar to the home-based than to the school-based workforce. For instance, 17.1 percent of school-based workers have at most a high-school degree. This proportion ascends to 39.8 percent and 50.7 percent among center- and home-based workers, respectively. Similarly, whereas school-based workers earn an average annual income of $27,014, center workers earn on average just over half this amount ($14,567) and the annual earnings of home-based workers are even lower ($12,415). Finally, while 13.6 percent of those who were school-based ECCE workers in 2009 had left the ECCE industry by 2010 the industry turnover rate among center- and home-based workers in 2010 was 24.4 and 28.5 percent, respectively. Table 2. Evolution of the ECCE Workforce by Sector (1990–2010) Center-BasedHome-BasedSchool-Based WorkersWorkersWorkers 199220101992201019922010 Distribution of the workforce by educational attainment Less than high school 12.3 9.8 37.6 19.8* 5.3 5.1 High school degree 32.7 30.0 34.5 30.9 20.6 12.0* Some college or associate's degree 33.3 36.6 21.8 34.3* 17.5 21.7 At least a bachelor's degree 21.6 23.7 6.1 15.0* 56.6 61.2 1990 2009 1990 2009 1990 2009 Mean annual earnings of all worker 10,809 14,567* 6,480 12,415* 24,191 27,014 (at 2010 dollars) Mean hourly earnings of full-year workers 9.2 10.9* 5.6 8.9* 17.5 18.2 (at 2010 dollars) Share of workers with pension and/or 20.4 24.5 3.1 7.6* 64.3 68.8 health benefits paid at least partly by the employer 1990 2010 1990 2010 1990 2010 Industry turnover rate 34.0 24.4* 36.9 28.5* 15.9 13.6 Average occupational prestige in the year 41.3 44.6 32.3 33.4 51.4 54.1 before entering the ECCE workforce Center-BasedHome-BasedSchool-Based WorkersWorkersWorkers 199220101992201019922010 Distribution of the workforce by educational attainment Less than high school 12.3 9.8 37.6 19.8* 5.3 5.1 High school degree 32.7 30.0 34.5 30.9 20.6 12.0* Some college or associate's degree 33.3 36.6 21.8 34.3* 17.5 21.7 At least a bachelor's degree 21.6 23.7 6.1 15.0* 56.6 61.2 1990 2009 1990 2009 1990 2009 Mean annual earnings of all worker 10,809 14,567* 6,480 12,415* 24,191 27,014 (at 2010 dollars) Mean hourly earnings of full-year workers 9.2 10.9* 5.6 8.9* 17.5 18.2 (at 2010 dollars) Share of workers with pension and/or 20.4 24.5 3.1 7.6* 64.3 68.8 health benefits paid at least partly by the employer 1990 2010 1990 2010 1990 2010 Industry turnover rate 34.0 24.4* 36.9 28.5* 15.9 13.6 Average occupational prestige in the year 41.3 44.6 32.3 33.4 51.4 54.1 before entering the ECCE workforce Notes: *Denotes change with respect to 1990 or 1992 is statistically significantly different from zero at the 5% level. Source: Authors based on the March Supplement of the Current Population Survey. ### Changes in the Characteristics of the ECCE Workforce in 1990–2010 The very low levels of formal education, compensation, and stability among the ECCE workforce warrant concern. As table 1 indicates, however, we also find meaningful signs of improvement. In fact, among the ECCE workforce as a whole, we show that all of the characteristics analyzed—education, compensation, turnover, and prestige of entrants—exhibited significant and substantial changes in the direction hypothesized to improve ECCE quality.11 As shown in table 1, the share of ECCE workers with at least some college education rose from 47 to 62 percent between 1992 and 2010. Mean annual earnings increased by 51 percent, from$10,746 to $16,215 between 1990 and 2009. Although part of this increase was driven by an increase in the number of hours worked by ECCE workers,12 the mean hourly earnings of ECCE workers also increased substantially over that period (by 33 percent, from$8.80 to \$11.70 per hour), and so did the share of ECCE workers with employer-paid pension and/or health benefits (from 19 to 28 percent). Annual turnover from the ECCE industry decreased substantially over the period of analysis (from 32.9 percent in 1990 to 23.6 percent in 2010). Finally, individuals who moved into child care from other occupations in 2010 came from somewhat more prestigious occupations than those who moved into child care in 1990. The average occupational prestige score of ECCE entrants increased by 4.7 percentile points over this period, from 37.6 to 42.3, perhaps indicating an improvement in the ECCE industry's ability to attract more qualified workers.

The changes observed among the ECCE workforce do not simply reflect trends in the female labor force and/or in low-wage industries. Compared with female workers, the ECCE workforce exhibited a larger increase in compensation and a steeper decline in industry turnover; and compared with low-wage workers, all variables exhibited a larger improvement among ECCE workers. Further, the changes observed reflect a stable trend within the industry and are not the product of the economic crisis that began in 2008.13

### Sector-Specific Changes?

In table 2 we show that the overall improvements seen in this workforce are driven by improvements among home-based workers, and, to a lesser extent, center-based workers.

In the home-based sector, the average educational attainment, compensation, and industry turnover of workers improved significantly and substantially over the period of analysis. With respect to educational attainment, there was a significant increase in the share of workers with at least some college (by 21.4 percentage points [p.p.]), and a significant decrease in the share of workers with less than a high school degree (by 17.8 p.p.). The average annual and hourly earnings of home-based workers increased by 92 and 50 percent, respectively, and the share of home-based workers with pension or health benefits rose as well (by 4.5 p.p.). Finally, industry turnover declined among home-based workers (by 8.4 p.p., from 36.9 percent in 1990 to 28.5 in 2010).

Changes within the center-based sector also suggest improvements over time, but these changes are of a smaller magnitude. For instance, between 1990 and 2009, the average annual earnings among center-based ECCE workers increased by 35 percent and average hourly earnings rose by 18 percent. Industry turnover rate dropped significantly, from 34 percent in 1990 to 24.4 percent in 2010. Other characteristics appear to change in a direction consistent with improvement, although the changes are not statistically significant. Differences remain between the sectors with respect to all the characteristics analyzed, but the pronounced changes within the home-based sector imply a narrowing of the gap with respect to the other two sectors.

### Expansion of Formal Care as an Explanation for Gains?

As shown in the fourth panel of figure 1, between 1990 and 2010 there was a significant change in the relative importance of the ECCE sectors in accounting for the size of the aggregate workforce. The relative importance of home-based workers declined sharply (by 21.8 p.p.), compensated mostly by an increase in the relative importance of center-based workers (by 17.5 p.p.). Although the relative importance of school-based workers increased only slightly (by 4.3 p.p.), the number of workers in this sector increased by 45 percent over this time period, a trend consistent with both the expansion of state pre-kindergarten programs and the shift toward full-day kindergartens. The number of center-based workers also increased dramatically (by 61 percent), while the number of home-based workers decreased (by 39 percent). This redistribution of ECCE workers from childcare homes to centers and schools is consistent with the recent decline in the share of children under age 5 years whose main childcare arrangement is in a home setting (U.S. Census Bureau 2010).

Figure 1.

Evolution of Selected Characteristics of the ECCE Workforce, and of the Relative Importance of Each ECCE Sector, Over Time (1990–2010)

Figure 1.

Evolution of Selected Characteristics of the ECCE Workforce, and of the Relative Importance of Each ECCE Sector, Over Time (1990–2010)

As discussed earlier, home-based workers have far lower levels of education and compensation and higher levels of industry turnover than do center- or school-based workers. The decline in the relative importance of home-based workers is one plausible explanation for the observed increase in the educational attainment, compensation, and stability of the national ECCE workforce. Changes in these characteristics within sectors are also relevant, however—and, in fact, more relevant than the changes in the distribution of the workforce across sectors.

We decompose aggregate changes in the ECCE workforce into the part explained by the expansion of the formal sector and the part explained by changes in the characteristics of workers within the sectors. We present the estimations in panel A of table 3. Although both factors contribute to the overall change, for most variables (educational attainment, annual and hourly wages, and industry turnover), changes within the sectors explain most of the aggregate improvement, with changes in the relative importance of the sectors explaining only a small portion of the overall improvement. For example, increases in earnings within sectors explain 78 percent of the overall increase in annual earnings, whereas the redistribution of workers across sectors explains only 22 percent. Similarly, within-sector changes explain 86 percent of the decline in industry turnover.

Table 3.
Decomposition of the Overall Changes in the Characteristics of the ECCE Workforce (1990–2010)
Panel APanel B
Change attributable toChange attributable toSector contributions to the part of the
changes in the characteristicschanges in the distributionchange attributable to changes in the
of workers within the sectorsof workers across sectorscharacteristics of workers within the sectors
Center-BasedHome-BasedSchool-Based
WorkersWorkersWorkers
2010 vs. 19922010 vs. 1992
Distribution of the workforce by educational attainment
Less than high school −8.8 (65%) −4.7 (35%) 12% 88% 0%
High school degree −4.0 (84%) −0.8 (16%) 28% 39% 32%
Some college or associate's degree 7.4 (84%) 1.4 (16%) 19% 73% 8%
At least a bachelor's degree 5.4 (58%) 3.9 (42%) 16% 71% 13%
2009 vs. 1990 2009 vs. 1990
Mean annual earnings of all workers (at 2010 dollars) 42% (78%) 12% (22%) 37% 55% 8%
Mean hourly earnings of full-year workers 25% (72%) 10% (28%) 35% 60% 4%
(at 2010 dollors)
Share of workers with pension and/or health benefits 4.3 (48%) 4.7 (52%) 41% 45% 14%
paid at least partly by the employer
2010 vs. 1990 2010 vs. 1990
Industry turnover rate −8.1 (86%) −1.3 (14%) 53% 44% 4%
Average occupational prestige in the year before 2.1 (49%) 2.1 (51%) 58% 28% 14%
entering the ECCE workforce
Panel APanel B
Change attributable toChange attributable toSector contributions to the part of the
changes in the characteristicschanges in the distributionchange attributable to changes in the
of workers within the sectorsof workers across sectorscharacteristics of workers within the sectors
Center-BasedHome-BasedSchool-Based
WorkersWorkersWorkers
2010 vs. 19922010 vs. 1992
Distribution of the workforce by educational attainment
Less than high school −8.8 (65%) −4.7 (35%) 12% 88% 0%
High school degree −4.0 (84%) −0.8 (16%) 28% 39% 32%
Some college or associate's degree 7.4 (84%) 1.4 (16%) 19% 73% 8%
At least a bachelor's degree 5.4 (58%) 3.9 (42%) 16% 71% 13%
2009 vs. 1990 2009 vs. 1990
Mean annual earnings of all workers (at 2010 dollars) 42% (78%) 12% (22%) 37% 55% 8%
Mean hourly earnings of full-year workers 25% (72%) 10% (28%) 35% 60% 4%
(at 2010 dollors)
Share of workers with pension and/or health benefits 4.3 (48%) 4.7 (52%) 41% 45% 14%
paid at least partly by the employer
2010 vs. 1990 2010 vs. 1990
Industry turnover rate −8.1 (86%) −1.3 (14%) 53% 44% 4%
Average occupational prestige in the year before 2.1 (49%) 2.1 (51%) 58% 28% 14%
entering the ECCE workforce

Notes: Changes in the share of workers by educational attainment, the share with pension and/or health benefits, and the industry turnover rate are measured in percentage points; changes in annual and hourly earnings, as a percent change; and change in the average occupational prestige score of those entering the ECCE workforce, in percentiles.

Source: Authors based on the March Supplement of the Current Population Survey.

Further, as reported in panel B of table 3, changes within the home-based workforce explain most of the change in educational attainment and earnings that is attributable to within-sector changes. Indeed, improvements within the home-based sector drive over two thirds of the increases in the ECCE workforce's educational attainment.

## Discussion

This policy brief highlights the current state of the ECCE workforce and explores whether this workforce has experienced meaningful changes over a period characterized by heightened interest and investment in early childhood programs. Echoing earlier work, we find that this workforce continues to be characterized by very low levels of education, compensation, and stability. However, we also show that both the educational attainment and the compensation of the ECCE workforce increased meaningfully between 1990 and 2010, and turnover from the ECCE industry decreased substantially. Taken together the findings are mixed, highlighting both improvements over time and the persistence of troubling issues. For example, our data show that in 1992, ECCE workers with a BA earned 47 percent less than elementary school teachers with the same educational level. Despite the significant increases in both educational attainment and earnings among ECCE workers that we document in this paper, in 2009 ECCE workers still earned 38 percent less than elementary school teachers. Given our increased understanding of the importance of early childhood interventions and of high-quality ECCE providers, these patterns are concerning. The positive trends we document, however, suggest that substantial changes in this workforce are, in fact, taking place.

It is worth noting that the positive trends we document differ significantly from those reported in prior studies, which document a decline or modest change in the educational attainment and compensation of the ECCE workforce. One explanation is that prior studies have generally focused on the center-based workforce and have not accounted for the evolution of the home-based workforce, where we find meaningful improvements (Whitebook et al. 2001; Bellm and Whitebook 2006; Herzenberg, Price, and Bradley 2005; Saluja, Early, and Clifford 2002).

A second explanation is that our study makes use of more current data than earlier work. For instance, an earlier study that relies on the same data used here but tracks the center-based workforce only through 2003 reports a decline in the proportion of that workforce that holds a BA (Herzenberg, Price, and Bradley 2005). We replicate that finding here, but show that between 2004 and 2010 this trend is reversed. Overall, we do not observe significant changes (either increases or decreases) in the educational attainment of the center-based workforce over the period 1990–2010, but document significant improvements in the compensation and stability of this workforce.

We also document a dramatic reconfiguration of the ECCE workforce, such that the majority of workers now work in formal rather than home-based settings. Surprisingly, however, we show the shift away from home-based care and toward center-based settings is not the primary explanation for the improvements observed in the industry at large. In fact, most of the improvements in the ECCE workforce are explained by within-sector improvements in the characteristics of workers. Further, whereas the center-based workforce exhibited significant increases in earnings and a remarkable decline in industry turnover, improvements within the home-based workforce were the primary driver of the increase in the educational attainment and earnings of the aggregate ECCE workforce.

These findings—that the overall improvement of the ECCE workforce was primarily driven by improvements within the home-based workforce—are surprising in light of the policy emphasis on expanding and improving formalized ECCE settings, such as preschools and pre-kindergarten programs, over informal settings. Improvements within the home-based workforce may be the result of recent efforts to increase the qualifications and stability of these workers. For instance, recent initiatives reward participation in professional development and the acquisition of further education, supplement the wages of home-based workers to ensure they meet a locally determined minimum living wage, and facilitate the provision of employer-sponsored health plans by pooling together workers from different childcare centers and homes (Kagan, Kauerz, and Tarrant 2007). Still, further study is needed to understand what has driven the observed improvement in the education, compensation, and stability of home-based workers, and to understand how to continue this positive and unexpected trend.

### Study Limitations

Although the current study provides new evidence about the current status of the ECCE workforce and its changing nature over the past two decades, the CPS was not designed to study the ECCE industry and several of its limitations are worth highlighting.

First, the CPS, while commonly used in analyses of workers, relies on self-reported data. To the extent that certain segments of the ECCE workforce are less likely to report their employment, our estimates will not accurately generalize to the ECCE workforce in its entirety. Further, if these nonreporters have lower earnings and educational attainment than do other workers, our findings will overestimate conditions in this industry, a troubling point given the already low levels we document. Although we are unable to assess the extent of nonreporting in our sample, it is likely we exclude some portion of the informal sector including unpaid workers, paid workers who do not report taxes, or paid family, friends, and neighbors who, despite assuming childcare responsibilities, do not report it as a job. These informal settings represent a meaningful portion of the market, and more nuanced data are necessary to better understand the composition of this group.

Second, the CPS does not provide direct measures of care quality and thus cannot be used to assess whether and how more proximal measures of care quality have changed. Although our outcomes provide a clear picture of the economic status of the ECCE workforce, an important issue in its own right, ultimately policy makers wish to improve early childhood experiences for children and the relationship between each of these measures and care quality is not as well understood as we would like. It is difficult to know, for example, to what extent changes in earnings over time amount to better care for young children. We have interpreted our findings as indicative of improvements in the quality of the ECCE workforce, but a competing hypothesis is that the increase in ECCE workers’ compensation and the reduction of turnover reflect an increase in the demand for ECCE services without a corresponding improvement in the actual quality of these workers. Additional research investigating the link between structural measures such as the ones available in administrative data sets would help here.

Third, our statistical inferences are limited by our small sample size. Each March, the CPS surveys around 670 center-based workers, 530 home-based workers, and 230 school-based ECCE workers. Using three-year moving averages, we were able to describe the evolution of the center- and home-based workforces with reasonable precision. However, our sample size was too small to make reliable inferences about the evolution of the school-based workforce.

Finally, the CPS cannot be used to distinguish between ECCE workers who work with infants and toddlers, and those who work with preschoolers. Similarly, we are unable to distinguish between pre-kindergarten and kindergarten employees. Data that allow for these types of delineations would better allow us to unpack trends and begin to understand the mechanisms driving these patterns.

## Conclusion

Although our findings echo other recent work on the low levels of earnings and education within the ECCE workforce, this brief also sheds an optimistic light on the possibility for improvements. We show that the qualifications, compensation, and stability of the ECCE workforce can improve, and in fact have improved meaningfully over the past two decades. The decline in turnover from the ECCE industry has been particularly marked. While some degree of turnover may be desirable in order to replace ineffective workers, the annual ECCE industry turnover rate in 1990 was 32.9 percent, roughly three times higher than the industry turnover rate of 11 percent observed among elementary and secondary education teachers. By 2010, however, the gap between the two had narrowed significantly, owing to the reduction in turnover among ECCE workers. To our knowledge, ours is the first study to look at the evolution of turnover for a nationally representative sample of the ECCE workforce. Although we are unable to observe job turnover, which is a more proximal measure of the instability children experience, industry turnover is an important measure in its own right, showing that individuals are staying within the industry longer than they did in the past, which may translate to positive outcomes for children and may indicate that early childhood jobs are more attractive than they once were.

The improvements we have identified for ECCE workers have taken place within both the center- and home-based sectors, which together account for over 80 percent of the workforce. Improvements within home-based child care have been particularly pronounced. To the extent that the characteristics we analyzed are, in fact, proxies of ECCE quality, our findings imply a narrowing in the quality gap between home-based and other more formalized types of child care. This finding is important because as recently as 2005, the home-based sector, historically singled out as the lowest-quality sector within child care, served around 40 percent of children under 5 years whose mothers were employed (U.S. Census Bureau 2010), and there is some evidence that it is the preferred type of arrangement among Hispanic families (Fuller, Holloway, and Liang 1996; Liang, Fuller, and Singer 2000; Fuller 2007). Put differently, workers in childcare homes remain substantially less qualified than workers in the formal childcare sector. On the other hand, the trends we observe suggest some closing of the quality gap between sectors is possible.

## Notes

1

Education and turnover statistics present a similar picture. For instance, turnover in California child care centers between 1996 and 2000 was estimated at about 75 percent (Whitebook et al. 2001) and another study surveying child care centers in Iowa, Kansas, Nebraska, and Missouri found that 40 percent of caregivers intended to leave the ECCE industry within less than five years (Torquati, Raikes, and Huddleston-Casas 2007).

2

Although ideally we could also assess changes over time in direct measures of caregiver quality, national data tracking such measures over time are not available. Several studies have sought to determine whether there is a causal relationship between our proxies and the quality of care children experience. The evidence here is mixed. As described earlier, improvements in teachers’ educational attainment are often pursued as a strategy to improve quality, and some studies suggest that, over some range, higher levels of education are related to better classroom practices (Blau 2000). On the other hand, Early et al. (2007) raise doubts about the relationship between specific degrees and child outcomes. Higher wages are associated with better classroom practices and lower turnover from ECCE jobs (Blau 2000; Whitebook and Sakai 2003). Although we are not aware of studies investigating the impact of industry turnover on children's development, the few studies on the role of job turnover show that children who spend more time with their caregiver, and those who do not experience a change in the primary caregiver over the course of a year, establish more nurturing relationships with their caregiver and exhibit better cognitive outcomes (Elicker, Fortner-Wood, and Noppe 1999; Tran and Winsler 2011).

3

On average over the period 1990–2010, 82.8 percent of individuals identified as center-based ECCE workers were employed in the “child day care services” industry; the remaining 17.2 percent were in other industries.

4

Our “home-based workforce” includes all individuals who take care of a relative, friend, or neighbor's child, who report this to be their job. The CPS relies on self-reports and some relatives, friends, and neighbors who assume child care responsibilities may not report this as a job and will therefore be excluded from our analysis. To the extent that those who fail to report their employment may differ in important ways from those who do identify this way, our characterization may suffer from bias.

5

Information on educational attainment is available from 1992 to 2010. Most other workforce characteristics are available for 1990 to 2010. The exception is information on earnings and benefits available from 1990 to 2009.

6

We make this restriction because the CPS collects information about hourly wages only for a subsample of the March interviewees that excludes all self-employed individuals, thus excluding a large proportion of home-based workers. Rather than excluding home-based workers in our analysis, we estimated hourly wages of ECCE workers based on their annual earnings and their reported hours worked in a typical week. Because the CPS does not specify the number of weeks worked in the past year, we limited analysis to full-year workers for whom we assumed 50 weeks of work (available in a separate online technical appendix that can be found on the Education Finance and Policy Web site at www.mitpressjournals.org/efp). Note that our estimates therefore apply only to those ECCE workers who were employed on a full-year basis (i.e., those who worked nine months or more). These represent 46 and 65 percent of those workers who in 1990 and 2010, respectively, reported that their main job in the previous year had been an ECCE job. The subset of full-year ECCE workers appears to be slightly more educated than the aggregate ECCE workforce, although the differences between the two groups are not statistically significant. Still, our estimation may overestimate the hourly earnings of the aggregate workforce.

7

We limit this analysis to those workers with one job to isolate whether or not ECCE jobs provide benefits. Among all workers whose main job in the previous calendar year was an ECCE job, the proportion who had only one employer increased from 75 percent in 1990 to 84 percent in 2010. In specification checks that include all ECCE workers irrespective of the number of jobs worked, results (available upon request) are virtually unchanged.

8

Female workers are a relevant comparison group because females constitute the vast majority of ECCE workers. Based on our calculations, over 95 percent of ECCE workers over the period of analysis were women. The low-wage worker comparison includes workers from the main industries from which ECCE workers come when they enter the child care industry, as well as to which ECCE workers migrate when they leave the ECCE workforce. We consider the following industries: beauty salons, food services, entertainment and recreation services, grocery stores, department stores, and non-teaching jobs in elementary and secondary schools (e.g., bus drivers, cooks, janitors, teacher aides, secretaries, and administrative assistants). Together, over the full period of the study, these industries represent about a third of migration from another industry into child care, and from child care to another industry.

9

First, we estimate what the overall change in the ECCE workforce's characteristics would have been had the distribution of the workforce across the three sectors (center, homes, and schools) changed as it did, but assuming that the characteristics of workers within each sector remained the same as in 1990. Then, to estimate the part of the overall change that is driven by changes in the characteristics of workers within each sector, we estimate what the overall change in the workforce's characteristics would have been had the characteristics of the workers within each of the sectors changed as they did, but assuming the distribution of the workforce across the sectors remained the same as in 1990. The equations used for these simulations are provided in the online appendix, which is available at www.mitpressjournals.org/doi/suppl/10.1162/EDFP_a_00114.

10

Recall also that these figures likely overestimate the true compensation of the full ECCE workforce, due to our sampling restrictions (e.g., hourly wages are calculated based on full-year workers, benefits are calculated based on workers with only one job in the past year).

11

The changes in educational attainment, compensation, and industry turnover that we discuss throughout are statistically significantly different from zero at the 5 percent level. Changes in the occupational prestige score of ECCE entrants are significantly different from zero at the 15 percent level. Note that the analysis of average occupational prestige scores applies only to individuals who entered the ECCE workforce in a given year. This is a very small sample, so we evaluate significance at the 5, 10, and 15 percent levels.

12

The mean hours worked per week increased from 29.9 to 31.8 between 1990 and 2010.

13

One plausible hypothesis is that the observed improvements in ECCE workers’ qualifications and stability are the product of the economic crisis. However, in supplementary analysis available upon request, we explored whether there were changes in trends following the economic crisis that began in 2008. We find no evidence to support this claim and, if anything, our results suggest that the improvement in ECCE workers’ characteristics was stalled or reversed during the crisis period.

## Acknowledgments

We are grateful to David Deming, Bruce Fuller, Deborah Stipek, and two anonymous referees for useful comments on previous drafts of this paper. This research was supported by a grant from the Institute of Education Sciences (R305A100574). The views expressed here are those of the authors and do not necessarily reflect the views of the University of Virginia, Cornell University, Stanford University, or IES. Any remaining errors are our own.

## REFERENCES

Barnett
,
W. Steven
,
Dale J.
Ackerman
,
Megan E.
Carolan
,
Jen
Fitzgerald
,
Debra J.
Ackerman
, and
Allison H.
Friedman
.
2010
.
The state of preschool: 2010 state preschool yearbook
.
New Brunswick, NJ
:
National Institute for Early Education Research, Rutgers University
.
Bassok
,
Daphna
.
2010
.
Do Black and Hispanic children benefit more from preschool? Understanding differences in preschool effects across racial groups.
Child Development
81
(
6
):
1828
45
. doi:10.1111/j.1467-8624.2010.01513.x
Bassok
,
Daphna
,
Maria
Fitzpatrick
,
Erica
Greenberg
, and
Susanna
Loeb
.
2012
.
How much do sectors differ? Comparing children's experiences in formal and informal care settings
.
Unpublished paper, University of Virginia
.
Bassok
,
Daphna
,
Maria
Fitzpatrick
, and
Susanna
Loeb
.
2012
.
Disparities in child care availability across communities: Differential reflection of targeted interventions and local demand
.
Unpublished paper, University of Virginia
.
Bellm
,
Dan
, and
Marcy
Whitebook
.
2006
.
Roots of decline: How government policy has de-educated teachers of young children
.
Available
www.plan4preschool.org/documents/roots-decline.pdf.
Accessed 4 December 2012
.
Blau
,
David M.
2000
.
The production of quality in child-care centers: Another look.
Applied Developmental Science
4
(
3
):
136
48
Brandon
,
Richard N.
, and
Ivelisse
Martinez-Beck
.
2006
.
Estimating the size and characteristics of the United States early care and education workforce
. In
Critical issues in early childhood professional development
, edited by
Martha Zaslow and Ivelisse Martinez-Beck
, pp.
49
70
.
Baltimore, MD
:
Paul H. Brookes Publishing Co
.
Early
,
Diane M.
,
Kelly L.
Maxwell
,
Margaret
Burchinal
,
Soumya
Alva
,
Randall H.
Bender
,
Donna
Bryant
,
Karen
Cai
, et al
2007
.
Teachers’ education, classroom quality, and young children's academic skills: Results from seven studies of preschool programs.
Child Development
78
(
2
):
558
80
. doi:10.1111/j.1467-8624.2007.01014.x
Elicker
,
James
,
Cheryl
Fortner-Wood
, and
Ilene C.
Noppe
.
1999
.
The context of infant attachment in family child care.
Journal of Applied Developmental Psychology
20
(
2
):
319
36
. doi:10.1016/S0193-3973(99)00019-2
Fox
,
Sharon E.
,
Pat
Levitt
, and
Charles A.
Nelson
.
2010
.
How the timing and quality of early experiences influence the development of brain architecture.
Child Development
81
(
1
):
28
40
. doi:10.1111/j.1467-8624.2009.01380.x
Fuller
,
Bruce
.
2007
.
Standardized childhood: The political and cultural struggle over early education
.
Stanford, CA
:
Stanford University Press
.
Fuller
,
Bruce
,
Susan
Holloway
, and
Xiaoyan
Liang
.
1996
.
Family selection of child-care centers: The influence of household support, ethnicity, and parental practices.
Child Development
67
(
6
):
3320
37
. doi:10.2307/1131781
Government Accountability Office (GAO)
.
2012
.
Early child care and education: HHS and education are taking steps to improve workforce data and enhance worker quality
.
Washington, DC
:
GAO
.
Hamre
,
Bridget
, and
Robert C.
Pianta
.
2006
.
Student–teacher relationships
. In
Children's needs III: Development, prevention, and intervention
, edited by
George G. Beer and Kathleen M. Minke
, pp.
59
72
.
Washington, DC
:
National Association of School Psychologists
.
Harris
,
Douglas N.
, and
Scott J.
.
2007
.
Understanding the level and causes of teacher turnover: A comparison with other professions.
Economics of Education Review
26
(
3
):
325
37
. doi:10.1016/j.econedurev.2005.09.007
Helburn
,
Suzanne W.
1995
.
Cost, quality and child outcomes in child care centers. Technical report, public report, and executive summary
, vol.
1–3
.
Denver, CO
:
.
Herzenberg
,
Stephen
,
Mark
Price
, and
David
.
2005
.
Losing ground in early childhood education: Declining workforce qualifications in an expanding industry, 1979–2004
.
Washington, DC
:
Economic Policy Institute
.
Howes
,
Carollee
,
Deborah A.
Phillips
, and
Marcy
Whitebook
.
1992
.
Threshold of quality: Implications for the social development of children in center-based child care.
Child Development
63
(
2
):
449
60
.
Institute of Medicine (IOM) and National Research Council (NRC)
.
2012
.
The early childhood care and education workforce: Challenges and opportunities: A workshop report
.
Washington, DC
:
.
Kagan
,
Sharon Lynn
,
Kristie
Kauerz
, and
Kate
Tarrant
.
2007
.
The early care and education teaching workforce at the fulcrum: An agenda for reform
.
New York
:
Teachers College Press
.
Knudsen
,
Eric I.
,
James J.
Heckman
,
Judy L.
Cameron
, and
Jack P.
Shonkoff
.
2006
.
Economic, neurobiological, and behavioral perspectives on building America's future workforce.
Proceedings of the National Academy of Sciences of the United States of America
103
(
27
):
10155
62
. doi:10.1073/pnas.0600888103
Liang
,
Xioyan
,
Bruce
Fuller
, and
Judy D.
Singer
.
2000
.
Ethnic differences in child care selection: The influence of family structure, parental practices, and home language.
Early Childhood Research Quarterly
15
(
3
):
357
84
. doi:10.1016/S0885-2006(00)00071-5
Nam
,
Charles B.
2000
.
Comparison of three occupational scales
.
Unpublished paper, Florida State University
.
Nam
,
Charles B.
, and
Monica
Boyd
.
2004
.
Occupational status in 2000: Over a century of census-based measurement.
Population Research and Policy Review
23
(
4
):
327
58
. doi:10.1023/B:POPU.0000040045.51228.34
National Institute of Child Health and Human Development (NICHD) Early Child Care Research Network
.
2000
.
Characteristics and quality of child care for toddlers and preschoolers.
Applied Developmental Science
4
(
3
):
116
35
Peisner-Feinberg
,
Ellen S.
,
Margaret R.
Burchinal
,
Richard M.
Clifford
,
Mary L.
Culkin
,
Carollee
Howes
,
Sharon Lynn
Kagan
, and
Noreen
Yazejian
.
2001
.
The relation of preschool child-care quality to children's cognitive and social developmental trajectories through second grade.
Child Development
72
(
5
):
1534
53
. doi:10.1111/1467-8624.00364
Phillips
,
Deborah
, and
Marcy
Whitebook
.
1986
.
Who are child care workers? The search for answers.
Young Children
41
(
4
):
14
20
.
Saluja
,
Gitanjali
,
Diane M.
Early
, and
Richard M.
Clifford
.
2002
.
Demographic characteristics of early childhood teachers and structural elements of early care and education in the United States.
Early Childhood Research & Practice
4
(
1
):
1
20
.
Shonkoff
,
Jack P.
, and
Deborah
Phillips
.
2000
.
From neurons to neighborhoods: The science of early childhood development
.
Washington, DC
:
.
Torquati
,
Julia C.
,
Helen
Raikes
, and
Catherine A.
Huddleston-Casas
.
2007
.
Teacher education, motivation, compensation, workplace support, and links to quality of center-based child care and teachers’ intention to stay in the early childhood profession.
Early Childhood Research Quarterly
22
(
2
):
261
75
. doi:10.1016/j.ecresq.2007.03.004
Tout
,
Kathryn
,
Rebecca
Starr
,
Margaret
Soli
,
Shannon
Moodie
,
Gretchen
Kirby
, and
Kimberly
Boller
.
2010
.
The child care quality rating system (QRS) assessment: Compendium of quality rating systems and evaluations
.
Washington, DC
:
Child Trends and Mathematica Policy Research
.
Tran
,
Henry
, and
Winsler
.
2011
.
Teacher and center stability and school readiness among low-income, ethnically diverse children in subsidized, center-based child care.
Children and Youth Services Review
33
(
11
):
2241
52
. doi:10.1016/j.childyouth.2011.07.008
U.S. Census Bureau
.
2010
.
Historical and time series table: Primary child care arrangements of preschoolers with employed mothers: Selected years, 1985 to 2010
.
Available
www.census.gov/hhes/childcare/data/sipp/index.html.
Accessed 21 January 2013
.
U.S. Department of Education (USDOE)
.
2011
.
Race to the top—Early learning challenge: Guidance and frequently asked questions for applicants and grantees
.
Available
www2.ed.gov/programs/racetothetop-earlylearningchallenge/faq.html (
FAQs combined). Accessed 21 January 2013
.
Vandell
,
Deborah Lowe
, and
Barbara
Wolfe
.
2000
.
Child care quality: Does it matter and does it need to be improved? Special report No. 78
.
:
Institute for Research on Poverty
.
Whitebook
,
Marcy
, and
Laura
Sakai
.
2003
.
Turnover begets turnover: An examination of job and occupational instability among child care center staff.
Early Childhood Research Quarterly
18
(
3
):
273
93
. doi:10.1016/S0885-2006(03)00040-1
Whitebook
,
Marcy
,
Laura
Sakai
,
Emily
Gerber
, and
Carollee
Howes
.
2001
.
Then and now: Changes in child care staffing, 1994–2000. Technical Report
.
Washington, DC
:
Center for the Child Workforce
.