Donate
Independent, objective, nonpartisan research

R 1203MOR

Authors

R 1203MOR

Tagged with:

Publication PDFs

Database

This is the content currently stored in the post and postmeta tables.

View live version

object(Timber\Post)#3711 (44) { ["ImageClass"]=> string(12) "Timber\Image" ["PostClass"]=> string(11) "Timber\Post" ["TermClass"]=> string(11) "Timber\Term" ["object_type"]=> string(4) "post" ["custom"]=> array(5) { ["_wp_attached_file"]=> string(13) "R_1203MOR.pdf" ["wpmf_size"]=> string(6) "643748" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(141401) "Arranging and Paying for Child Care ••• Margaret O’Brien-Strain Laura Moyé Freya Lund Sonenstein 2003 PUBLIC POLICY INSTITUTE OF CALIFORNIA Library of Congress Cataloging-in-Publication Data O’Brien-Strain, Margaret. Arranging and paying for child care / Margaret O’Brien-Strain, Laura Moyé, Freya Lund Sonenstein. p. cm. Includes bibliographical references. ISBN: 1-58213-047-7 1. Child care services—California. 2. Preschool children— Care—California. I. Moyé, Laura. II. Sonenstein, Freya L. III. Title. HQ778.65.C2O27 2003 362.71'2'09794—dc22 2003023308 Copyright © 2003 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. PPIC does not take or support positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword This report by Margaret O’Brien Strain, Laura Moyé, and Freya Lund Sonenstein provides a useful and timely analysis of child care for young children in California. What share of families use child care, and what type of care? How much public assistance do they receive to secure the care? How does California’s experience compare with that in the rest of the nation? How expensive would a universal preschool program be, and what are the alternatives to such a program? These are all questions central to any state governor facing growing budget deficits, and especially to a governor with an avowed interest in the future of the state as manifested in its children. The authors conclude that California families are less likely than families elsewhere in the United States to use child care of any kind for their young children. This is especially true for low-income families and those with less-educated parents. Children age 4 in California are also much less likely than those in the rest of the nation to attend preschool and, not surprisingly, this is especially true for children in lower-income families. Finally, the authors estimate that a full-year, full-day universal preschool care program would cost as much as $5 billion annually. Given that such a program is unlikely to emerge from the current budget crisis, or anything approaching a major increase in child care expenditures, the authors suggest that other options are likely to be on the table—for example, encouraging kindergarten enrollment for lowincome children, supported by Head Start. All options will be considered when welfare reform legislation is reauthorized in Washington in the next year or two. Encouraging parents to leave welfare is one thing. Finding the ways and means to support child iii care assistance to help working families is yet another. It is obvious that the first strategy cannot be successful without some substantial help from the other. David W. Lyon President and CEO Public Policy Institute of California iv Summary California’s child care policies represent a complex set of programs attempting to address the goals of both supporting working parents and promoting child development. As California faces a budget crisis and a major political change, these policies may be subject to a variety of changes. The 2003–2004 state budget included some reductions to the system of child care subsidies, which at its peak provided over $1.5 billion to help working families pay for child care, especially families transitioning from welfare to work. Governor Davis proposed much more drastic changes to the subsidy programs, but the most serious review of these policies is expected to follow the delayed federal reauthorization of welfare reform. Current federal proposals call for much more stringent work requirements, and the level of federal support for child care has been one of the stumbling blocks for passage of reauthorization legislation. At the same time, California has a number of programs aimed more toward promoting early childhood education, including state-funded preschools, contracted slots in child development programs, services to increase the quality of care, and resource and referral services. Although programs such as Head Start are built on the recognition that children at risk for school failure could benefit greatly from early childhood programs, recent pressures to improve school outcomes have turned public attention to the role of early childhood programs in preparing all children for school. Unlike Head Start or state-subsidized preschools that target groups based on income, universal preschool initiatives seek to promote access to quality early childhood programs for all preschool-age children on a voluntary basis. Thus, the California Master Plan for Education calls for universal preschool for children ages 3 and 4, a concept embraced by the David and Lucile Packard Foundation, First 5 California, and several county First 5 commissions, including Los v Angeles and San Mateo, which are all supporting exploratory efforts to implement such a program. In light of potential changes in child care policy, this report examines the current state of child care for young children in California. In particular, we address four key questions: • What share of California families use what types of child care for their young children, and how does this differ from the rest of the United States? • How much of their household income do families pay for child care and how much assistance do they receive in paying for it? • Are there major gaps between California and the rest of the country in preschool enrollment? If so, how much is explained by demographic differences? • If preschool access is expanded through universal preschool programs, how expensive will these programs be and how do different designs affect the cost of care through such programs? The analysis combines data on families with children ages 0–5 in the 1997 and 1999 rounds of the National Survey of America’s Families, for a total of 1,287 children ages 0–5 from 1,282 California families. We consider a broad range of care settings regularly used by parents of young children: structured care in programs such as child care centers, nursery schools, preschools, and Head Start programs; family day care provided to a group of children in the provider’s home; care by relatives in the child’s home or the relative’s home; and nanny or babysitter arrangements where an unrelated caregiver regularly provides child care in the child’s home. We examine all regular arrangements used by both working and non-working parents. Our key findings are listed below. • California families are less likely to use child care for their young children than are families elsewhere in the United States. In part, this is due to lower labor force participation by California parents and a larger share of families in groups that are lower users of child care. Even within these groups, however, child care use is lower in California than in the rest of the United States. Still, most children of working parents (83 percent) vi regularly spend time in non-parental care, averaging 35 hours per week. Forty-five percent of children with a parent not in the labor force are also regularly placed in non-parental care, albeit for fewer hours per week than the children of working parents. • Children with less-educated parents and poorer children are less likely to be placed in care. Because their families are disproportionately poor and less educated, Hispanic children in particular are less likely to be placed in care. Higher-income families are more likely than lower-income families to place children in structured care settings, although families formerly on welfare have patterns of child care use very similar to those of higher-income families. • Children of working single parents spend the most time in care and are also the most likely to be in multiple care arrangements. Although multiple care arrangements are common even for children with parents out of the labor force, the ability to coordinate work schedules has a strong influence on the number of arrangements used for young children, as well as on their hours in care. Families with some non-standard work hours can limit both the number of hours in care and the number of arrangements. Such families have greater reliance on care provided by relatives when they do use child care. • Families who pay for child care while parents are working pay an average of $373 per month out-of-pocket, representing about 10 percent of family earnings. This cost is higher than the average in the rest of the United States but approximately the same share of income. Poor families who do pay for care pay an average of 24 percent of earnings, compared to about 7 percent for families with income above the state median. About 28 percent of all working parents pay no out-of-pocket costs for child care. • Most families at or below 75 percent of the state median income— the eligibility cutoff for child care subsidies, about $38,000—receive some assistance in paying for child care, although often as free care provided by relatives. Forty-six percent of poor families, onethird of former welfare recipients, and two-thirds of current recipients receive government or social services assistance in vii paying for care. However, current welfare recipients are less likely to receive free care from relatives, perhaps because the welfare system will reimburse relatives for providing care. Use of structured care is much higher among welfare recipients and other low-income families receiving assistance. • Children age 4 in California are much less likely than those in the rest of the nation to attend preschool. Only 50 percent of these children in California are enrolled in preschool, compared to 63 percent in the rest of the United States. A similar gap holds for children age 3, but California’s relatively early kindergarten enrollment age makes up for the U.S.-California gap in preschool attendance for children age 5. • As with child care use overall, children with less-educated mothers or from lower-income families are less likely to attend preschool. Families with more familial support, from stay-at-home mothers, other adults in the home, or other available relatives, are also less likely to use child care. Together these factors help create a particular gap in preschool enrollment of Hispanic children: only 35 percent of Hispanic children ages 3 and 4 attend preschool compared to 45 percent of non-Hispanic children. • Full-year, full-day universal preschool care could cost as much as $5 billion annually when fully implemented, although if eligibility were limited to families with income below the cutoff for the current subsidy system, this cost would fall to between $1.4 billion and $2.7 billion. The costs of universal preschool vary widely depending on assumptions of eligibility criteria and on takeup of preschool by children not currently enrolled, but a substantial share of the costs would likely come from the shift of children currently enrolled in preschool from private payment to public funding. Part-day preschool would be substantially less expensive, around $1.3 billion to $2.3 billion for a universal program. Although part-time programs are much less accessible to working parents, the lower use of preschool by low-income families with a parent not in the workforce may make this limitation a secondary concern. Costs of implementing and administering the program are not included in these estimates. viii The findings in this report point to other options, short of universal preschool, for improving early childhood education. One important first step would be encouraging kindergarten enrollment. The same families least likely to use preschool are also disproportionately likely to have children age 5 not enrolled in kindergarten. This is particularly an issue for low-income and Hispanic families. Many children who would be eligible for programs such as Head Start are being cared for by their mothers or other relatives; outreach to and education of these groups could increase their involvement in the existing system. However, even with outreach, access to Head Start and state preschool programs would have to be addressed to fully accommodate eligible families. Finally, some families appear to select settings other than preschool because partday programs, such as those offered by the state preschools, do not match well with the needs of working parents. Nor should calls for universal preschool draw attention away from the ongoing need for child care assistance to help working parents. Reducing eligibility for subsidies will increase the costs of work for many families, increasing the share of income spent on child care. Although relatives caring for young children for free may absorb some of the costs, reduced subsidies may also force some parents currently using structured care settings to switch to more informal arrangements. If welfare reauthorization further increases the work requirements for welfare recipients, additional resources will be needed to assist families. ix Contents Foreword ..................................... Summary..................................... Figures ...................................... Tables ....................................... Acknowledgments ............................... iii v xiii xv xvii 1. INTRODUCTION ........................... 2. DATA .................................... The National Survey of America’s Families ............ Sample of Families with Young Children .............. Analysis Variables............................. Child Care Arrangements ...................... Child Care Expenses and Assistance ............... Economic and Demographic Variables .............. 3. OVERVIEW OF CHILD CARE USE IN CALIFORNIA .. Share of Children Under Age 6 in Child Care ........... Choice of Arrangements ........................ Choice of Arrangements, by Child and Family Characteristics ........................... Work Schedules and Care Arrangements .............. Summary .................................. 4. PAYING FOR CHILD CARE .................... Monthly Expenditures on Child Care ................ Monthly Expenditures, by Family and Child Characteristics ........................... Child Care Assistance .......................... Summary .................................. 5. PRESCHOOL ENROLLMENT .................. Care Settings for Preschool-Age Children ............. Characteristics Linked to Preschool Attendance.......... 1 5 5 6 8 8 9 10 13 13 14 18 23 27 29 29 30 32 36 39 40 43 xi Targeting Efforts to Groups Less Likely to Use Preschool ............................... Children in Parental Care Only .................. Children in Other Care Settings .................. Promoting Use of Preschool ...................... 6. COSTS OF UNIVERSAL PRESCHOOL............. Assumptions on Takeup of and Costs for Preschool ....... Families Eligible for Universal Preschool............. Enrollment of Children Not Currently in Preschool ..... Crowd-Out of Existing Preschool ................. Cost per Child in Universal Preschool .............. Alternative Estimates of the Costs of Universal Preschool ... Summary .................................. 7. CONCLUSIONS ............................ Appendix: Supplementary Tables ..................... Bibliography .................................. About the Authors ............................... Related PPIC Publications .......................... 50 50 52 53 55 55 56 57 57 59 60 64 65 67 73 75 77 xii Figures 3.1. Percentage of Children Ages 0–5 in Each Type of Care Arrangement ............................. 3.2. Percentage of Children Ages 0–5 in Each Type of Care Arrangement When All Parents Are Employed ....... 3.3. Percentage of California Children Ages 0–5 in Each Type of Care Arrangement, by Age ............... 4.1. Average Monthly Expenditure on Child Care by Parents Paying for Care ........................... 4.2. Payment Arrangements of Families Working and on Welfare ................................ 4.3. Use of Structured Care by California Families With and Without Child Care Assistance ................. 5.1. Child Care Settings for Children Ages 3–5 .......... 5.2a. Child Care Settings for Children Age 5 ............ 5.2b. Child Care Settings for Children Age 4 ............ 5.2c. Child Care Settings for Children Age 3 ............ 5.3. Share of California Children Ages 3 and 4 in Preschool, by Region ............................... 5.4. Share of California Children Ages 3 and 4 in Preschool, by Child Care Capacity in County of Residence ...... 5.5. Child Care Settings for Children Ages 3 and 4, by Household Composition/Employment ............ 5.6. Child Care Settings for Children Ages 3–5, by Number of Children Under Age 13 .................... 5.7. Child Care Settings for Children Ages 3–5 Not in Preschool ............................... 16 17 18 30 34 36 41 42 42 43 46 46 49 51 52 xiii Tables 2.1. Sample Characteristics of Children Ages 0–5 and Their Families ................................ 3.1. Percentage of Children Ages 0–5 in Child Care ....... 3.2. Percentage of California Children Ages 0–5 in Child Care, by Demographic Group and Type of Arrangement ............................. 3.3. Percentage of California Children Ages 0–5 in Child Care, by Work Schedule and Type of Arrangement .... 3.4. Hours and Number of Arrangements Used for California Children Ages 0–5 in Child Care, by Work Schedule ... 4.1. Monthly Child Care Expenditures for California Families Who Pay for Care, All Parents Employed ..... 4.2. Child Care Assistance for California Families Using Care for Children Ages 0–5, All Parents Employed ........ 5.1. Percentage of Children Ages 3 and 4 in Preschool, by Demographic Group ........................ 5.2. Regression Results on the Likelihood of California Children Ages 3 and 4 Being Enrolled in Preschool .... 5.3. Share of Preschool Arrangements That Are Part-, Moderate-, and Full-Time .................... 6.1. Assumptions on Takeup of Universal Preschool for California Children ......................... 6.2. Estimates of Average Annual Costs for Universal Preschool: High Estimate..................... 6.3. Estimates of Average Annual Costs for Universal Preschool: Low Estimate ..................... 6.4. Percentage of Costs Resulting from Participation of Currently Enrolled California Preschoolers, by Takeup Assumptions ............................. 6.5. Additional Slots Needed for California Children Enrolling in Preschool, by Takeup Assumptions ...... 7 14 20 24 25 31 33 45 48 53 59 60 61 63 63 xv A.1. Sample Characteristics for Family-Level Data (Costs) ... A.2. Percentage of Children Ages 0–5 in Rest of United States in Child Care, by Demographic Group and Type of Arrangement ....................... A.3. Counties Included in Each Region ............... A.4. Regression Results on Likelihood of Preschool Chldren Ages 3 and 4 Being Enrolled in California and Rest of United States ............................. A.5. Regression Results on Likelihood of Children Ages 3 and 4 Being Enrolled in Preschool in California ....... 67 68 70 71 72 xvi Acknowledgments This report would not have been possible without the extensive work of the Urban Institute’s Assessing the New Federalism project, their National Survey of America’s Families, and the assistance of their staff. Thanks as well to those who provided valuable comments including Deborah Reed, Shelley Waters Boots, Heather Rose, Susanna Loeb, and Laura Hill. Any opinions or errors are the authors’ alone. xvii 1. Introduction Child care policy in California reflects a struggle between two goals: supporting working parents and promoting child development. On the work support side, the 2000–2001 state budget included over $1.5 billion earmarked to help working families pay for child care, with the vast majority of funding going to families transitioning from welfare to work. The subsidies provided through this system reimburse parents for care provided by all but the most expensive providers. On the child development side, similar funding was provided through the Child Development Division of the California Department of Education (CDE) to support a variety of initiatives to promote access to highquality care, including support for state preschools, contracted slots in child development programs, services to increase the quality of care, and resource and referral services. Federal funding supports the Head Start program, which provides developmentally rich care for at-risk preschool children. As resources shrink in the state budget crisis, these two goals may come increasingly into conflict. The 2003–2004 budget made some cuts to child care subsidies; Governor Davis proposed a number of more dramatic changes, but they did not receive legislative support. These included deeper reductions in reimbursement rates, increases in family fees, and giving higher priority on waiting lists to families working fulltime. Other proposals called for moving control of child care subsidies from CDE to the Department of Social Services (DSS), which is more likely to focus on child care as a work support, or to the counties, where county flexibility could open the door to wide differences in county strategies to balance quality and affordability but could also put child care funding into competition with other county social service activities. Although such proposals seem to move the balance toward work support, other trends move it toward expansions in developmentally rich care, especially for children ages 3 and 4. Funded through a cigarette tax, 1 the California Children and Families Commission, also known as First 5 California, has made child development for children ages 0–5 one of its central goals. Two of its current initiatives alone provide over $500 million for school readiness programs and child development teacher retention. The Master Plan for Education, the David and Lucile Packard Foundation, and several county First 5 commissions, including Los Angeles and San Mateo, have embraced the call for universal preschool programs to increase child development and school readiness of children ages 3 and 4. In light of this rapidly changing policy environment, this report examines the current state of child care for young children in California, including the care arrangements currently used by families and the costs of this care. We then look more closely at issues from each side of the policy debate: first, the role of child care subsidies and other assistance in making care more affordable and, second, the issues involved in universal preschool. In particular, we address four key questions: • What share of California families use what types of child care for their young children, and how does this differ from the rest of the United States? • How much of their household income do families pay for child care and how much assistance do they receive in paying for it? • Are there major gaps between California and the rest of the country in preschool enrollment? If so, how much is explained by demographic differences? • If preschool access is expanded through universal preschool programs, how expensive will these programs be and how do different designs affect the cost of care through such programs? The analysis relies on 1997 and 1999 household survey data collected by the Urban Institute through the National Survey of America’s Families (NSAF). We consider a broad range of care settings: structured care in programs such as child care centers, nursery schools, preschools, and Head Start programs; family day care provided to a group of children in the provider’s home; care by relatives in the child’s home or the relative’s home; and nanny or babysitter arrangements 2 where an unrelated caregiver regularly provides child care in the child’s home. The report is organized as follows: Chapter 2 describes the National Survey of America’s Families, which provides the data for the report. Chapter 3 gives an overview of the level of child care use and the types of settings chosen by families with different demographic characteristics. Chapter 4 looks at the costs of care and the role of child care assistance, with some perspectives on the possible effect of proposals to change subsidy policies. Chapter 5 assesses preschool enrollments and examines the issues involved in promoting greater enrollment. Chapter 6 then addresses the effects of different assumptions on the estimated cost of universal preschool. We offer conclusions in Chapter 7. 3 2. Data The National Survey of America’s Families The data for this study are drawn from the National Survey of America’s Families, a multiyear survey conducted by the Urban Institute as part of its Assessing the New Federalism project.1 The survey was designed to learn more about the economic, health, and social circumstances of children and adults under age 65; it includes a representative sample of the U.S. non-institutionalized, civilian population with oversampling of low-income families, as well as additional observations in 13 target states, including California. Our study uses data from the first two survey rounds of the NSAF: 1997 and 1999. In 1997, 2,543 households in California were interviewed out of 44,000 households nationwide. The 1999 survey includes 2,206 California households out of over 42,000 interviewed nationwide.2 Interviews were conducted by telephone, and to capture households without telephones, interviews were conducted using cellular phones loaned to these households. In households with children under age 18, up to two children could be chosen for in-depth study: one child under age 6 and one child age 6–17. Interviews about these children were conducted with the adult in the household who was the most knowledgeable about each sampled child. In California, 95 percent of these adults were the biological parents. Eighty-one percent were female, of which most (91 percent) ____________ 1Information about Assessing the New Federalism can be found on the Urban Institute website at www.urban.org. 2Additional information on the design of the NSAF can be found on the Urban Institute website at http://newfederalism.urban.org/nsaf/methodology1997.html for 1997 and http://newfederalism.urban.org/nsaf/methodology1999.html for 1999. 5 were the biological mothers. Nineteen percent were male, of which 93 percent were the biological fathers.3 Sample of Families with Young Children These analyses of child care focus on families with a child age 0–5 that were interviewed during the school year in 1997 and 1999.4 Further, we compare eligible families living in California to similar families living in the rest of the United States. To retain an adequate number of individuals in our sample for California for this comparison, we combined the interviews conducted in 1997 and 1999. After combining the observations, our sample for California includes 1,287 children in 1,282 families. For the rest of the United States, our sample includes 20,350 children ages 0–5 in 20,328 families. Sample weights were adjusted to account for the different sample and population sizes in the two years.5 Families with young children in California differ substantially from families with young children in the rest of the nation. Table 2.1 presents an overview of the characteristics of the children and their parents. In particular, almost half of California children ages 0–5 are Hispanic compared to only 13 percent nationwide. California children are less likely to be African American and more likely to be Asian. However, the sample sizes for each of these two groups are too small to support subgroup analysis of the California sample. Therefore, we will focus only on differences between non-Hispanic whites and Hispanics in any breakdowns by race/ethnicity. Racial and ethnic makeup is not the only significant difference between California families and those living elsewhere in the nation. ____________ 3For simplicity, we will use the term parent(s) henceforth. 4Child care arrangements vary substantially between the summer months and other seasons (Cappiziano, Adelman, and Stagner, 2002). Thus, we focus our analyses on care occurring in non-summer months. The sample size is too small to allow examination of summer child care use in California. 5These weights adjust the sample weights developed by the Urban Institute to balance across the two years and scale up to an average of the population between 1997 and 1999. Discussion on the creation of the underlying NSAF weights can be found in the methodology reports at http://newfederalism.urban.org/nsaf/methodology.html. 6 Table 2.1 Sample Characteristics of Children Ages 0–5 and Their Families Characteristics of Children and Parents Child’s age 0–2 3–5 Child’s race/ethnicity Non-Hispanic white Hispanic Other Income as a % of federal poverty limit Less than 100 100–199 Over 200 Most knowledgeable adult’s welfare history Currently on welfare Formerly on welfare Never on welfare Most knowledgeable adult’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Most knowledgeable adult’s marital status Single parent Married or cohabiting parent Adults 18 or older in household in addition to parents Other adults in household No other adults in household California Weighted No. % 576 49 711 51 443 38 652 45 192 17 361 27 343 22 583 51 156 12 154 11 970 77 270 21 358 26 296 22 350 31 309 21 978 79 297 22 990 78 Rest of United States Weighted No. % 9,156 11,194 49 51 13,498 2,985 3,867 67 13 20 4,320 5,383 10,647 21 24 56 1,230 3,048 15,976 6 13 81 2,231 6,402 4,734 6,870 11 31 23 35 5,169 15,181 22 78 3,229 17,121 14 86 Children under age 6 living in families in California tend to be poorer than similar families in the rest of the United States. The most knowledgeable adults (typically mothers) interviewed about these children are also less educated, with twice as many not completing high school (21 percent compared to 11 percent). In the 1997–1999 period, welfare participation is also higher in California. Almost twice as many young children in California live with adults who are currently on 7 welfare (12 percent compared to 6 percent). Family structure for these children differs somewhat in California. Although single parenthood is about equally common in California as in the rest of the county, California children are more likely to live in extended families. Analysis Variables Child Care Arrangements The telephone interview included questions about the regular use of specific types of child care for the selected child during the past month. Regular use was defined as at least once a week during the past month. In addition, the adult respondents were asked a number of questions about the care used, including the number of hours the arrangements were used, whether the caretaker was a relative, and whether the arrangement was used while the respondent, usually the parent, worked, went to school, or participated in job training. Adult respondents were also asked whether the selected child attended kindergarten.6 From this information, we determine whether the selected child regularly participated in non-parental care in any of the following settings: center care, Head Start, nursery school, kindergarten, before and after school care, family day care, nanny or babysitter care, in-home care by a relative, and out-of-home care by a relative. When parents used none of these non-parental care arrangements, the child’s care setting is designated as parental care. In some cases, parents may take their children to work or work back-to-back schedules so that they can care for their children themselves. In other cases, the respondents may use irregular arrangements that vary from week to week or they may leave their children unattended but are reluctant to admit this in the interview. These cases are included in the parental care category, although the extent of their prevalence is unknown. We first constructed a four-category employment variable by household composition: single parent, employed; two parents, both employed; two parents, one employed; single or two parents, no one ____________ 6Because our data did not include birth dates, we cannot say how many of the children age 5 in our sample were eligible for kindergarten. 8 employed.7 We then divided households into two employment categories: fully employed and not fully employed. A household is fully employed if all parents are in the workforce regardless of the number of hours worked. A household is considered not fully employed if at least one parent is not in the workforce. Child Care Expenses and Assistance The NSAF asked parents how much they paid out-of-pocket for child care arrangements8 in the last month for all their children under age 13 while they worked, were in school, or looked for work.9 Thus, our analyses of child care expenses and the receipt of child care assistance refer to the costs incurred by the families with working parents for child care for all their children under age 13.10 We focus our analyses only on families who had at least one child age 0–5 and who used at least one child care arrangement while a parent was employed. Note that some families who had at least one non-employed parent also reported using non-parental child care arrangements. However, the survey did not collect information about the costs incurred for this care. We coded working parent families as receiving assistance in paying for child care based on their responses to a number of questions posed by the survey. All families that used child care for employment purposes were asked if anyone helped to pay for all or part of the cost of care, regardless of whether they reported any child care expenses. If they answered yes, they were asked to indicate who or what agency helped them pay for care. We focused our analyses on reports of assistance from ____________ 7Two-parent households include married parents as well as a parent and a cohabiting partner. 8It is not possible to examine the costs of each child care arrangement separately or to examine the aggregate costs of each child in the family’s care arrangement. 9If the two most knowledgeable adults lived in the same family (for example, each spouse or unmarried partner was most knowledgeable about a different focus child), the cost data each reported were combined and the family was treated as one case (Giannarelli, Adelman, and Schmidt, 2003). 10Appendix Table A.1 provides a breakdown of family characteristics in our cost sample. 9 three general sources: help from the government, help from a relative, and other help.11 In addition, if families did not report receiving help but they indicated that the amount they paid for care was determined by how much they earned, we coded them as receiving help from the government.12 Families that used child care while they worked and did not report receiving help but did not report any child care expenses were assumed to be receiving “free” child care. They were assigned to one of three categories of assistance depending on their child care arrangement. If they used structured care, we assumed that they were receiving free government help. If a relative provided child care, we assumed that they were receiving free relative care. And if they were in some other type of arrangement, such as using a nanny/babysitter or family day care, we assumed that they were receiving free other care. Cases that were in more than one type of arrangement were counted in more than one assistance category. For example, a family that reported no child care costs but that used a child care center part-time and care by a relative part-time would be included in both the free government help and free relative care categories. Economic and Demographic Variables Throughout this analysis, we compare the child care arrangement and cost patterns for different groups of children and families based on their economic and demographic characteristics. Although most of the categories are straightforward, several are explained below. ____________ 11Our analyses of child care assistance follow the Urban Institute’s approach to differentiating the sources of assistance (see Giannarelli, Adelman, and Schmidt, 2003). Some families who were designated as receiving “other” assistance may in fact be receiving a government subsidy. We cannot distinguish with certainty the sources of assistance for families who are in a private child care program and receiving assistance. The subsidies that some of these families are receiving may be funded through a government program. For this reason, government assistance may be underestimated (Giannarelli, Adelman, and Scmidt, 2003). 12Some of these families could in fact be paying on a sliding scale with a private organization that is not reimbursed by the government for this assistance; therefore, “other” assistance may be underestimated. 10 In addition to poverty status, we categorized a family’s annual income relative to the State Median Income (SMI). For all families, we divided their income by California’s median income for their respective family size and interview year. We chose to use this categorization because income relative to the SMI determines which families are eligible for subsidized child care in California, our focus state. We constructed a measure of welfare history using survey data on current and past participation in welfare (Temporary Assistance for Needy Families, TANF, or Aid to Families with Dependent Children, AFDC). Using these data and the SMI ratio, we constructed a welfare history variable with four categories: (1) currently on welfare, (2) formerly on welfare (this included families that had ever received welfare), (3) never received welfare but income-eligible for subsidies, and (4) higher income and never received welfare.13 ____________ 13The third and fourth groups are distinguished by whether their income was above or below 75 percent of California SMI, the eligibility cutoff at the time of the surveys. 11 3. Overview of Child Care Use in California The complexity of child care policy arises from the complexity of child care itself. Parents use non-parental care to allow them to work and to provide a socially and developmentally enriching environment for children, especially as children approach school age. In this chapter, we provide an overview of the use of child care in California and in the rest of the nation, including the arrangements chosen by parents—ranging from formal Head Start programs to informal care provided by relatives in the children’s homes—and the characteristics of children and families linked to different choices of care. Share of Children Under Age 6 in Child Care Sixty-five percent of California children under age 6 were regularly placed in at least one child care arrangement in a typical month between 1997 and 1999. As shown in Table 3.1, this rate was significantly lower than the rate of child care use for children in the rest of the nation, where 73 percent were in care. This difference is partly driven by the lower labor force participation of parents of young children in California. Fifty-four percent of U.S. children live in households where all parents are employed, compared to only 46 percent in California, primarily because of lower labor force participation of mothers in two-parent households. (Labor force participation among single parents is also lower in California.) However, child care use is lower in California regardless of employment status or household structure. We distinguish between single-parent and two-parent families and between families in which all parents are employed or not (both employed in the case of two-parent families or one employed in the case of one-parent families). Single parenthood is about as likely in California as in the rest of the 13 Table 3.1 Percentage of Children Ages 0–5 in Child Care All children Parent(s) employed Single parent, employed Two parents, both employed Parent(s) not employed Single parent, not employed Two parents, one or both not employed California 100 46* 10* 36* 54* 11* 43* *Indicates significance at the 95 percent level. Rest of United States 100 54 14 41 46 8 38 % of Group in Non-Parental Care Rest of United California States 65* 73 83* 88 89 92 81 87 49* 54 62 64 45* 52 United States, but single parents are less likely to work in California. Similarly, two-parent families in California are less likely to have both adults in the labor force. Child care use is highest among working single parents and lowest among two-parent families with one or both parents out of the labor force. For each of the four family types, use of child care is lower in California than in the rest of the United States, although within these groups the California-U.S. difference in care use is significant only for two-parent families with one or both adults not employed. Although child care is frequently viewed as a support for working parents, the majority of young children in the United States spend time in non-parental care even when there is at least one parent not working, in school, or in training. The share is lower in California, but even in California nearly half of young children with non-working parents are in at least one care arrangement. In large part, this reflects the importance of child care arrangements in supporting child development, an issue we return to in Chapter 5. Choice of Arrangements Child care is offered in a number of different settings. The grouping of these arrangements depends on the question to be considered. In this 14 section, we are interested in the options selected by families—the result of decisions balancing their preferred modes of care, the care available in their community, the costs of various settings, and the convenience and quality of the alternatives. In considering the options parents selected, we group settings into five major categories: Structured care represents the most formal arrangement, including care at both centers and schools. This care is most likely to be licensed and, typically, the most consciously oriented toward child development. Under this category, we include child care centers, nursery schools, preschools, and Head Start programs. In addition, some children age 5 are enrolled in kindergarten.1 Although kindergarten may be thought of as school rather than child care, it is optional in California. The state also allows younger children into kindergarten than do many other states. Because California families may have greater opportunities to substitute kindergarten for other developmentally rich child care arrangements, we include kindergarten in the structured care category.2 Family day care is care provided to a group of children in the provider’s home. These providers may or may not be licensed, depending on state regulations. In California, providers who care for children from more than one family (other than their own) must be licensed. However, the NSAF data do not distinguish between licensed and unlicensed family day care. Relative care is divided between care provided by relatives in the child’s home and care provided by relatives in the relative’s home. Extended family members, such as grandparents, have always been an important and trusted source of care for young children. Such care when provided in the child’s home is often given by extended family living in the same household, whereas such care outside the child’s home could ____________ 1Surveys were completed between February and November, so some children age 4 could have been eligible for kindergarten at the time of the fall surveys. However, we did not observe any of them in kindergarten. 2Earlier Urban Institute analyses of care options excluded kindergartners. However, given that center care is relatively less common for children under age 6 in California, and kindergarten attendance is somewhat more common, we believe that the inclusion of kindergartners in structured care more accurately reflects the greater use of such care in California than in the rest of the United States. 15 range from a very informal arrangement with a grandmother to a businesslike arrangement that differs little from family day care. Thus, parents may or may not have to pay for care given by a relative. Nanny or babysitter care represents the final category in the analysis below. These are arrangements where an unrelated caregiver provides child care in the child’s home.3 We do not distinguish between nannies and babysitters, although only regular ongoing arrangements are included. Figure 3.1 shows the share of young children cared for regularly in each of these five categories.4 70 60 *58 *52 50 California Rest of the United States Percentage 40 30 22 21 25 25 †27 †23 20 14 12 10 0 Structured care Family day care Care by Care by relative Nanny/ relative in outside babysitter child’s home child’s home NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. † Indicates significance at the 90 percent level. Figure 3.1—Percentage of Children Ages 0–5 in Each Type of Care Arrangement ____________ 3Nanny care offered in another person’s home is generally not distinguishable from family day care. 4This categorization of care settings differs from that reported in Urban Institute publications in two ways. The Urban Institute typically reports only the primary care setting (the setting in which a child spends the most hours) and concentrates on care used while parents are employed. 16 Structured care is the most common setting for children under age 6. In California, 52 percent of young children in some care arrangement spend time in a child care center, nursery school, preschool, Head Start, or kindergarten during a normal week. Care by relatives is the next most common setting. In-home and out-of-home care by relatives combined accounts for nearly as many children as structured care. California families do not differ substantially from other American families in the type of child care used, except that young children in California are less likely to be in structured settings. Care by relatives outside the home is also somewhat less common for California children, and care by nannies or babysitters is slightly more common. The share of children cared for in family day care homes, in California and elsewhere, rises if we restrict our view to families where all parents are employed (Figure 3.2). This shift to family day care is somewhat more pronounced for California children. 70 60 *57 50 *50 California Rest of the United States Percentage 40 30 27 24 *24 *26 25 28 20 15 12 10 0 Structured care Family day care Care by Care by relative Nanny/ relative in outside babysitter child’s home child’s home NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 3.2—Percentage of Children Ages 0–5 in Each Type of Care Arrangement When All Parents Are Employed 17 Choice of Arrangements, by Child and Family Characteristics Parents’ choice of arrangements for their young children is a complex decision, but clearly characteristics both of the children and their families are important. The age of the child is perhaps the most important characteristic for understanding the choice of arrangements. Parents of very young children, those under age 3, are less likely to place their children in care. Even though 65 percent of all young California children are in some care arrangement, only 52 percent of children ages 0–2 are in care, compared to 76 percent of children ages 3–5. One in five infants and toddlers with working parents are still cared for by parents only, meaning that work schedules must be arranged to keep a parent at home with the child. In contrast, structured care in California is primarily used for children ages 3–5, as we see in Figure 3.3. Without conditioning on a child being in care, infants and toddlers are more likely than older children to be cared Percentage 90 80 *76 70 60 50 40 30 20 *19 10 0 Structured care Children Ages 0–2 Children Ages 3–5 *37 *18 *31 *18 *25 *18 17 12 Care by Care by relative relative in outside childs home child’s home Family day care Nanny/ babysitter NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 3.3—Percentage of California Children Ages 0–5 in Each Type of Care Arrangement, by Age 18 for by relatives. Among children in care, all arrangements other than structured care are more common for the youngest children. The age of the child, however, is not the only demographic characteristic that is important in understanding choice of care arrangements. Table 3.2 shows the choice of arrangements for young children in California disaggregated by six family characteristics: family structure, child’s race/ethnicity, parental education, the presence of other adults in the household, income, and welfare receipt.5 Income and education are both important factors in characterizing which families use child care. The first column of Table 3.2 shows the share of children in each group in at least one non-parental arrangement, regardless of parents’ employment status. Disadvantaged families are the least likely to use any child care. Families with income below poverty are least likely to use non-parental care. As income rises, the share of families using care also rises. The share using child care also rises with education. Fifty-five percent of parents with less than a high school diploma use child care for their young children, compared to 73 percent of parents with at least an associate degree. Families with less income and less education also have lower labor force participation, but the differences in child care use hold even if we consider only working families. One of the most striking differences in child care use is for Hispanic children. Forty-one percent of these children are cared for by parents only, compared to 29 percent of non-Hispanic white children. However, this difference can be largely explained by other factors. Hispanic children are more likely to live in two-parent families, with lower average income, lower average education, and larger family sizes. Controlling for these characteristics, there is no statistically significant gap between Hispanic and non-Hispanic families in the likelihood of using child care for their young children. Hispanic families who do use child care for their young children are less likely to select structured care settings and more likely to select relatives as caretakers. Not surprisingly, the presence of other adults in the household has a large influence on the care choices of parents. Fortyseven percent of young children with another adult in the home are cared ____________ 5Appendix Table A.2 shows a similar breakdown for the rest of the United States. 19 Table 3.2 Percentage of California Children Ages 0–5 in Child Care, by Demographic Group and Type of Arrangement % in NonParental Care % in Structured Care by Relative Care in Child’s Home % in Arrangementa Care by Relative Outside Child’s Family Home Day Care Nanny/ Babysitter 20 Family structure Single parent Two parents Child’s race/ethnicityb Non-Hispanic white Hispanic Parental educationc Less than high school High school diploma/GED Some vocational education/college Associate degree or more Other adults in household No Yes Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI 75* 62 71 59* 55* 58 69* 73* 65 64 52* 65* 70 73 60 32 51 23 56 18 46* 34* 47 33 48 28 54 26 59* 21 54 18 47* 47* 55 25 47* 32* 51 28 58 19 20 22 8* 24 21 15 20 25 20 27* 17 8* 24 13 7 28 21 5 26 26 14* 18 23 21* 25 23 15 16 16* 3* 21 18 11* 28 18* 10* 23 22 2* 21 26 21 21 Table 3.2 (continued) Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % in NonParental Care % in Structured Care by Relative Care in Child’s Home % in Arrangementa Care by Relative Outside Child’s Family Home Day Care 72* 56 21 55 49 33 73* 55 21 59* 49 32 21 25 25 16 27 22 19 17 NOTE: Italics indicate reference group for significance tests. aChildren may be in more than one arrangement. bSample sizes in California are too small to report breakouts for black and Asian children. cBased on years of education of the “most knowledgeable adult” for the focal child. *Indicates significance at the 95 percent level. Nanny/ Babysitter 17 11 8 8 for by a relative in the home, compared to 18 percent of children without other adults in the home. The households of Hispanic children are much more likely to include other adults (27 percent, compared to 18 percent of households of non-Hispanic children), explaining the high use of relatives in the home by this group.6 Hispanic children, however, are also more likely to be cared for by a relative outside the home. The choice of child care arrangement significantly differs by parents’ level of education: Parents with more education are more likely to use structured care and nannies or babysitters. Similarly, an increase in income does not simply predict choice of care arrangement. The income categories in Table 3.2 are based on two conceptual threshold levels of income: the poverty threshold and the SMI. At present, families with income below 75 percent of the SMI are eligible for subsidized child care in California. The SMI measure in effect from September 1998, created by the California Department of Finance, places this median at $50,148 for a family of four. This means that the 75 percent cutoff fell at approximately $37,611—about 200 percent of the current federal poverty threshold of $18,400.7 As noted above, lower-income families are less likely to have their children in care, but there is no arrangement whose use rises consistently with income. For example, poor families are about as likely to use structured care as families with income above the state median income. Programs targeted to the poor may explain much of this similarity. Families just above the poverty threshold are less likely to use structured care. The opposite pattern holds for care by relatives: families with income low enough to qualify for subsidies but above the poverty threshold are more likely to rely on relatives than either higher- or lowerincome families. Families with income above the state median are most likely to have nannies or babysitters, but families in the next highest income category are the least likely. ____________ 6Other research on ethnic differences in child care selection found that Hispanic parents were significantly less likely than white parents to choose center care; however, after controlling for the presence of other adults in the home, the lower propensity of Hispanics to use center care diminished (Liang, Fuller, and Singer, 2000). 7The SMI cutoff has not been updated for inflation. 22 Welfare receipt has a curious connection to child care use. The last section of Table 3.2 disaggregates families into four groups based on welfare history and eligibility for subsidies: never received welfare and income too high to qualify for subsidies, never received welfare and income low enough to qualify for subsidies, did not receive welfare at the time of the survey but previously received it, and receiving welfare at the time of the survey. Former welfare recipients have both child care usage and arrangement types that are very similar to higher-income families. They have high child care usage, are most likely to use structured care, and are somewhat more likely to use family day care. Former welfare families, who typically transitioned from welfare to work, are eligible for child care assistance. Former welfare recipients’ higher use of licensed care and particularly center care is consistent with findings from administrative data on child care subsidies and surveys of subsidized child care settings (Marrufo, O’Brien-Strain, and Oliver, 2003).8 On the other hand, families currently on welfare have patterns very similar to those of low-income families who have never been on welfare. These families have lower use of any child care, and they are more likely than other families to rely on care by relatives, especially in the child’s home. Current welfare recipients also qualify for child care assistance, but only 26 percent of current recipient families have all parents working, compared to 56 percent of former welfare recipients. In the next chapter, we explore further the role of child care subsidies. Work Schedules and Care Arrangements About half of young children with a parent at home spend time regularly in a non-parental setting, as shown in Table 3.3. The majority of these children are placed in structured care settings, although other arrangements are also used, especially care provided by relatives. In families where all parents work (a working single parent or two parents who both work), work schedules influence both the probability of using non-parental care and, to a lesser extent, the type of arrangements chosen. Most notably, parents who work during evenings ____________ 8The higher use of structured care among former welfare recipients may result because former recipients switch their use of care as their work and household stabilize. 23 Table 3.3 Percentage of California Children Ages 0–5 in Child Care, by Work Schedule and Type of Arrangement 24 At least one parent not employed Parents employed At least one part-time, only regular hours Full-time, only regular hours At least one part-time, some evening or weekend hours Full-time, some evening or weekend hours % in NonParental Care 49* 83 87 61* 81 % in Structured Care 58 45 54 32 47 NOTE: Italics indicate reference group for significance tests. aChildren may be in more than one arrangement. *Indicates significant at the 95 percent level. % in Arrangementa Care by Relative Care by Relative Outside Child’s in Child’s Home Home 24 20 Family Day Care 13* Nanny/ Babysitter 11* 29 24 26 26 21 22 33 12* 37 24 16 28 33 32 17 11 or weekends are less likely to rely on non-parental care and most likely to use care by relatives. Use of structured care is not notably lower for these families, except for those working non-standard hours part-time. Babysitters and nannies are more commonly used by families where both parents are employed but not full-time. Reflecting the different uses of child care, the number of hours per week that children spend in care differs between families where all parents work and families with at least one parent at home. Not surprisingly, children with non-working parents spend much less time in care than do those with all parents in the workforce (Table 3.4). On average, a child with working parents spends 35 hours per week in care, whereas children with a parent at home spend 18 hours per week in care. Children whose parents work full-time during regular hours spend more time in non-parental care than do children whose parents work only part-time or non-standard schedules. This, coupled with the lower Table 3.4 Hours and Number of Arrangements Used for California Children Ages 0–5 in Child Care, by Work Schedule All children Children with At least one parent not employed Parents employed At least one part-time, only regular hours Full-time, only regular hours At least one part-time, some evening or weekend hours Full-time, some evening or weekend hours Parents employed Single parent Two parents Average Hours in All Arrangements 28 18* No. of Arrangementsa 3 or 1 2 More 60 32 8 66* 30 4 27 53 38 8 38* 54 34 12 23 60 33 7 31 60 30 11 41* 40* 43 17 32 61 31 9 NOTE: Italics indicate reference group for significance tests. aTotals may not sum to 100 because of rounding. *Indicates significance at the 95 percent level. 25 number of hours in care for children from two-parent families, demonstrates the role of reduced or offset schedules as a strategy for minimizing the time children spend in care. Children of working single parents spend an average of 41 hours per week in care, while those with two working parents spend only 32 hours per week in care on average. This difference results partly because of the shorter work hours common for second earners and partly because parents can vary their schedules to minimize the hours of child care needed. Across all these groups, the hours in care are virtually identical for California families and other American families. Overall, 40 percent of California children in care spend time in two or more different arrangements in an average week, as shown in Table 3.4. (Note that the figures in Table 3.2 are per child; thus, the complexity of care for families with multiple children is understated.) In some cases, the use of multiple arrangements appears to be a choice, probably related to a balance between developmentally rich care and affordable care, suggested by the fairly high use of multiple arrangements (34 percent) even among families with a parent not in the workforce. However, the need for multiple arrangements is also closely tied to work schedules: Working single parents, who lack a backup person to care for their children while they work, not only have their children in care for the longest hours but are also most likely to depend on multiple arrangements. This is particularly true of single parents in California, where 60 percent of single parents rely on multiple arrangements, compared to only 49 percent of single parents elsewhere in the United States. Although this information is not included in the table, the presence in the household of other adults, beyond the parents, counteracts this effect, where children with additional adults at home are least likely to be in multiple care arrangements when their parents work. Disadvantaged families are more likely to require multiple arrangements to facilitate work: Low-income working families and families combining welfare and work are the most likely to rely on multiple arrangements for very young children (under age 3). 26 Summary Fewer young children are placed in child care in California than elsewhere in the United States. Still, most children of working parents (83 percent) regularly spend time in non-parental care, averaging 35 hours per week. Children of working single parents spend the most time in care and are also the most likely to be in multiple care arrangements. However, child care is not used only to support work. Forty-five percent of children with a parent not in the labor force are also regularly placed in non-parental care, albeit for fewer hours per week than the children of working parents. Certain families are less likely to use care. In particular, children with less-educated parents and poorer children are less likely to be placed in care. Hispanic children are also less likely to be placed in care, although this difference is explained by the lower average education and lower average income for this group. Children under age 3 are less likely to be placed in care than are preschool-age children, and when they are in care, they are more likely to be cared for in an informal arrangement such as by relatives. Children with adults in the home other than their parents are also most likely to be cared for by relatives. Higher-income families are more likely than lower-income families to place children in structured care settings, although families formerly on welfare have patterns of child care use very similar to those of higher-income families. Finally, although multiple care arrangements are common even for children with parents out of the labor force, the ability to coordinate work schedules has a strong influence on the number of arrangements used for young children, as well as on their hours in care. Working single parents use both the most hours of care and the largest number of settings for their young children; families with some non-standard hours can limit both the number of hours in care and the number of arrangements. However, the coordination of part-time work and nonstandard hours seems to draw children away from structured care settings, with greater reliance on relatives when care is used. 27 4. Paying for Child Care Public policies concerned with child care typically focus on helping families pay for care. In California, child care assistance comes in a variety of forms, ranging from tax credits for all families—now available as a refundable credit—to targeted subsidies paid essentially as vouchers for current and former welfare recipients. In this chapter, we examine families’ expenditures for child care and the degree to which they receive assistance, from government or other sources, in paying for care. The NSAF captures information on the cost of child care in a different structure from the information it collects on child care arrangements. Whereas arrangement information is based on observations for a focal child, the cost of child care is captured at the family level for all children under age 13. Child care costs for all children in the family more accurately reflect the role of child care costs in a family budget and the effectiveness of assistance in supporting lowincome families. A major drawback is that we cannot relate costs directly to the care arrangements used. In addition, we cannot identify the specific arrangements for which families are receiving assistance. Monthly Expenditures on Child Care Among families who pay for child care, the average California family with young children spent $360 per month out-of-pocket on child care for all children in the family (Figure 4.1). As expected, given the differences in hours of care, families with at least one parent out of the labor force pay less for care, averaging $296.1 Families with both parents employed (or an employed single parent) pay $373 per month for care. Whether or not all parents are employed, California families on average pay more than families elsewhere for child care. However, because ____________ 1Families with no out-of-pocket costs are not included in the average cost calculations. 29 Expenditure ($) 400 350 300 250 200 150 100 50 0 All families who pay for care California Rest of United States Families with all parents employed Families with parents at home Figure 4.1—Average Monthly Expenditure on Child Care by Parents Paying for Care California families had higher earnings on average, child care costs were actually a slightly smaller share of earnings: 10 percent in California, compared to 11 percent elsewhere. Child care policies that help pay for care, as opposed to providing free care through Head Start or state preschools, typically require that parents be working to qualify for aid. This support is generally designed to make work more affordable, whether provided through the welfare system, the non-welfare subsidy system, or the tax system. For this reason, the rest of this chapter focuses on families with all parents employed. Monthly Expenditures, by Family and Child Characteristics Since monthly expenditures are reported by family rather than by child, we would expect child care expenditures to vary by the number and ages of children served. Indeed, families that include a child under age 3 typically had to spend about $36 more per month on child care than families in which the youngest child was age 3–5 (Table 4.1). However, families with two or more children under age 13 (one of 30 Table 4.1 Monthly Child Care Expenditures for California Families Who Pay for Care, All Parents Employed All working parent families Child’s race/ethnicity Non-Hispanic white Hispanic Other adults in household No Yes Age of youngest child 0–2 3–5 Number of children under age 13 1 2 or more Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % Paying for Care 72 Expenditures Mean Median As % of ($) ($) Earnings 373 320 10 75 430 400 11 69 273 240 11 77 383 325 10 55 324 311 13 71 392 332 11 74 356 320 10 73 365 325 10 71 379 311 11 57 242 194 24 67 283 300 13 75 350 268 9 73 417 400 7 76 426 400 7 66 262 249 13 69 309 260 13 59 414 337 29 whom is under age 6) do not pay significantly more than families with only one child requiring child care. This may reflect different strategies for families with several children or the presence of older siblings to help care for younger children. We do see that families with other adults in the household pay less for care, when they do pay for care, than families with no other adults. Hispanic families also pay much less for child care than do non-Hispanic white families, even after controlling for differences in income and the presence of other adults in the household. However, child care as a share of earnings is very similar for the two racial/ethnic groups. 31 Family income is the most important determinant of monthly expenditures on child care. Families with income above the state median paid an average of $417 per month on child care, compared to $242 paid by poor families. However, despite paying much less in absolute terms, lower-income families pay much more as a share of earnings. Poor families who do pay for care pay an average of 24 percent of earnings, compared to about 7 percent for families with income above the state median. Working families who are also receiving welfare pay the highest child care costs as a share of earnings. As we discuss below, most current welfare families receive assistance with child care, but among those who pay for care, the reported monthly expenditure of $414 on average is similar to that paid by high-income families. However, this appears to be driven by a few families paying very high amounts (and perhaps getting reimbursed), since the median monthly expenditure, although still higher than that for other low-income families, is much closer to that for the typical family.2 Child Care Assistance Many parents whose young children are in child care while the parents are working receive assistance in paying for this care (Table 4.2). Twenty-eight percent pay no out-of-pocket costs, because the care is provided free by relatives, subsidized by the government (through either subsidy vouchers, state supported preschools and kindergarten, the Head Start program, or other government or social service assistance), or paid by employers or absent parents.3 This is true even for higher-income families: 25 percent of families with income over 75 percent of the state median income paid no out-of-pocket expenses for child care. Families with other adults in the household were the least likely to pay for care. In California, only 55 percent of working families with other adults in the household paid for any care arrangements. One in ____________ 2It is possible that some parents included reimbursed cost when they reported their out-of-pocket costs. 3Families may also receive a dependent care tax credit, but these credits are not included in NSAF questions about child care assistance. 32 Table 4.2 Child Care Assistance for California Families Using Care for Children Ages 0–5, All Parents Employed All employed parent families Child’s race/ethnicity Non-Hispanic white Hispanic Other adults in household No Yes Age of youngest child 0–2 3–5 Number of children under age 13 1 2 or more Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % Paying No Child Care Costs 28 % Receiving Help Paying for Care Government Free Any or Social Care by Assistance Services Relative 37 19 14 25 31 31 44 12 12 26 14 23 32 45 56 17 11 26 24 29 33 26 41 15 13 23 14 27 37 29 36 43 74 33 46 25 43 27 28 17 16 21 12 46 14 28 10 18 13 11 12 24 25 34 51 31 54 41 68 8 14 27 16 33 14 66 6 four of these families received free care from relatives, nearly twice the rate of help from relatives for all working families in California. Free care by relatives was not the only source of assistance for families with other adults in the household: Nearly as many received subsidies or other assistance from the government or social service agencies.4 ____________ 4Care by relatives can be reimbursed through the welfare system or the non-welfare subsidy system. 33 One-third of families with income below 75 percent of the state median and 43 percent of poor families reported that they paid no outof-pocket expenses for child care. Almost three-quarters of poor families reported that they received some assistance in paying for care, including 46 percent receiving assistance from government or social services. This rate of government assistance is higher than the 36 percent rate for such families in the rest of the United States. More than one-fourth of California families with income above poverty but below 75 percent of the state median income also receive government help. This rate is much higher than elsewhere in the United States, because families in this range elsewhere are commonly above their local cap for federally funded assistance. The government assistance category, however, includes families that do not specifically report subsidies but are using free care through preschools or Head Start or other organized settings, so this may also partially reflect the continuing role of federal- and state-funded care in California. Welfare families are another group particularly likely to receive assistance with child care. In fact, in California, as in most states, working families on or recently off welfare are eligible for child care assistance through the TANF program. As Figure 4.2 shows, families 70 California 60 Rest of United States 50 Percentage 40 30 20 10 0 Pay for care Receive government Receive free care social services assistance from relative Figure 4.2—Payment Arrangements of Families Working and on Welfare 34 combining work and welfare in California are both more likely to receive government help paying for child care and more likely to pay some outof-pocket expenses. In fact, nationally only 36 percent of working families currently on welfare report paying for their child care, compared to 59 percent of those in California. But 66 percent of families report receiving government help in California, compared to 53 percent elsewhere. In California, in fact, it appears that there may be some tradeoff between government help and help from relatives, since only 6 percent of families currently on welfare report receiving free child care from relatives, less than one-third the share elsewhere. Surprisingly, child care assistance appears to often augment rather than offset spending on child care. Families with child care assistance are less likely to pay any out-of-pocket costs, but many low-income families receive help but still report paying for care. Families that pay for care while receiving assistance pay about the same amount per month as families with similar income but no assistance. For example, California families with working parents and income below 75 percent SMI paid on average $277 per month whether or not they received help with care. The effect is even more striking for welfare families: Those families currently on welfare and receiving assistance—when they pay for care— spend more than paying families without assistance: $460 per month compared to $395 per month. Because even among low-income families, those receiving help are likely to have lower earnings than those without help, families with assistance who still face costs are actually paying a larger share of their earnings for child care than families without assistance. In fact, low-income families may be using child care assistance not simply to lower the costs of care but also to allow them to select more developmentally oriented settings for their children. Among families with young children, low-income working parent families with assistance are far more likely to have their children in structured care settings than are such families without assistance.5 This is particularly true of families ____________ 5Research by Hirshberg, Huang, and Fuller (2002) of parents moving from cash aid to work activities in three California counties suggests that although subsidies may increase the likelihood of choosing center care, ethnicity, marital status, and home 35 currently receiving welfare: 68 percent of welfare families with assistance use structured care, compared to only 16 percent without assistance (Figure 4.3). On the other hand, the link between structured care and assistance may go the other way. That is, even though many forms of child care assistance, especially for welfare families, can be applied to all child care settings, it may be that assistance is much more closely tied to organized settings (through Head Start or state-subsidized preschools) or that families choosing organized care are much more likely to seek assistance than are families using other settings. 80 Families without assistance 70 Families with assistance 60 Percentage 50 40 30 20 10 0 Families with all parents employed Low-income families Families currently on welfare Figure 4.3—Use of Structured Care by California Families With and Without Child Care Assistance Summary Among families who use child care while both parents are working (or a single parent is working), 72 percent pay some out-of-pocket costs, averaging $373 per month. This cost is higher than the average in the rest of the United States but approximately the same share of income. However, there are significant disparities across income groups in the ________________________________________________________ language are likely to be part of the “causal process” that influences the selection of center care, “which in turn implies the use of a subsidized slot.” 36 share of income going to pay for child care. Poor families who do pay for care (57 percent of families using child care) pay an average of 24 percent of earnings, compared to about 7 percent for families with income above the state median. Twenty-eight percent of families receive free care, either because they receive government or social services assistance (including free care through Head Start) or because relatives provide free care to the children. Higher-income families pay more in absolute terms than do lower-income families, but lower-income families pay twice as large a share of their income for child care. Most families at or below 75 percent of the SMI receive some assistance in paying for child care. Forty-six percent of poor families receive government or social services assistance in paying for care. Onethird of former welfare recipients and two-thirds of current recipients who use child care while working receive government assistance in paying for care. However, current welfare recipients are less likely to receive free care from relatives, perhaps because the welfare system will reimburse relatives for providing care. However, use of structured care is much higher among welfare recipients and other low-income families receiving assistance. Reducing eligibility for subsidies will increase the costs of work for many families, increasing the share of income spent on child care. Although relatives caring for young children for free may absorb some of the costs, reduced subsidies may also force some parents currently using structured care settings to switch to more informal arrangements. 37 5. Preschool Enrollment Child care subsidies were dramatically increased in the late 1990s as an underpinning for welfare reform, but current policy trends concentrate on the role of child care as early childhood education. The foundational mission of First 5 programs in California, programs for children ages 0–5 financed through a 50 cent per pack cigarette tax approved under Proposition 10, is built on the conclusion that “current research in brain development clearly indicates that the emotional, physical and intellectual environment that a child is exposed to in the early years of life has a profound impact on how the brain is organized [and] . . . significantly influence[s] how a child will function in school and later in life.” Indeed research on early childhood intervention programs such as those reviewed by Karoly et al. (1998) and Gomby et al. (1995) found that intensive early childhood education programs had positive effects on cognitive and school outcomes, as well as social outcomes, with significant long-term cost savings. This work feeds into larger state policy concerns about school outcomes and school readiness from intensive early childhood education programs. As a result, one policy recommendation that has been gaining ground is a call for universal preschool. Voluntary universal preschool for children ages 3 and 4 was recommended as part of California’s new Master Plan for Education, because “the state has a profound interest in making available . . . the early education opportunities that support a child’s emotional, social, physical, linguistic, and cognitive development” (California State Senate, 2002). The Packard Foundation has made universal preschool a priority area for future grantmaking, Los Angeles County’s First 5 commission has dedicated an initial $100 million toward universal preschool in that county, and other county commissions are conducting feasibility studies for similar projects. As a background for understanding the potential desirability of universal preschool, this chapter examines more closely care settings for 39 preschool-age children. The following chapter presents a framework for estimating the costs of such proposals. Care Settings for Preschool-Age Children The call for universal preschool has particular relevance for California, where preschool-age children—those ages 3–5—are less likely than children elsewhere to be enrolled in a structured, center setting. In the previous chapter, we considered all types of care selected, recognizing that children were commonly cared for in multiple settings. To understand the issue of universal preschool, we need to consider whether children receive any school or center care.1 For this reason, we divide preschool-age children into four exclusive categories of care: • Not in care: Children cared for by their parents only. • Kindergarten: Among children in any care arrangement, we first distinguish children attending kindergarten. We classify them as kindergartners, although they may also receive care in other settings, including centers. • Preschool: Children who do not attend kindergarten but who do receive another type of structured care (preschools, nursery schools, Head Start programs, or child care centers) are classified as attending preschool. These children may also receive care in other settings. • Other care: Children in care who do not attend any kindergarten or preschool programs are classified as enrolled in other care settings. These children may be in more than one non-kindergarten, non-preschool arrangement. As Figure 5.1 shows, 57 percent of children ages 3–5 in California are enrolled either in preschool or in kindergarten, compared to 63 percent in the rest of the nation. More California children are enrolled in kindergarten—19 percent compared to 13 percent elsewhere—because ____________ 1Developmentally rich care may occur in non-center settings, and not all center programs are developmentally rich. However, most universal preschool/prekindergarten initiatives rely on center programs as well as schools. 40 Percentage 60 50 *50 40 *38 30 20 10 0 Preschool California Rest of United States *19 *13 19 20 *24 *17 Kindergarten Other care Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.1—Child Care Settings for Children Ages 3–5 California admits children at a younger age than the national average.2 However, a significant gap in preschool attendance leads to a large California-U.S. gap in enrollment in early childhood education. Enrollment in preschool and kindergarten varies greatly by age. Figure 5.2a, 5.2b, and 5.2c break down care settings by age within the preschool group. In our data, only children age 5 are enrolled in kindergarten, including 54 percent of children that age in California. Among children age 5, in fact, early enrollment in kindergarten brings the total share in structured settings in California above the average of the rest of the United States, with 88 percent in preschool or kindergarten, compared to 85 percent in the rest of the country. For children ages 3 and 4, however, the story is very different. There is a 13 percentage-point gap in preschool enrollment of children age 4 between California children and other U.S. children. There is a 10 percent gap for children age 3. ____________ 2California schools must admit children into kindergarten at the beginning of the school year if they will be five years of age on or before December 2 of the school year. Legislation passed in 2000 allows for pilot programs requiring that children be age 5 by September 1, but no such pilot programs have yet been funded. 41 Percentage 70 60 50 *46 40 *34 30 20 10 0 Preschool *54 *39 California Rest of United States Kindergarten 8 3 Other care *9 *7 Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.2a—Child Care Settings for Children Age 5 70 *63 60 50 *50 40 30 20 10 0 Preschool California Rest of United States 22 19 *28 *18 Other care Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.2b—Child Care Settings for Children Age 4 Percentage 42 Percentage 70 60 50 40 *38 30 *28 20 10 0 Preschool 35 35 California Rest of United States *36 *27 Other care Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.2c—Child Care Settings for Children Age 3 The difference in preschool enrollment is driven primarily by a larger share of California children not being placed in any non-parental care setting, rather than by greater use of other child care options, such as family day care, care by relatives, or care by a nanny or babysitter. For example, we find that California children age 3 were equally likely to be in other care settings but 9 percentage points more likely than other American children to be cared for only by their parents. Children not in any care setting also explain most of the difference for children age 4. Although children age 5 who are not yet in kindergarten would be a target group for universal preschool programs, it is easier to consider the issues for preschool attendance by focusing on children ages 3 and 4. For this reason, we will first look only at preschool attendance for these children, and then return to the issue of kindergarten later in the chapter. Characteristics Linked to Preschool Attendance The lower workforce participation of California parents partially explains the higher level of parental-only care in California, but preschool enrollment is lower in California regardless of employment status. Excluding children age 5, 53 percent of preschool-age children in 43 California have at least one parent not in the workforce, compared to 45 percent of children elsewhere. However, the gap in preschool enrollment is even larger for children ages 3 and 4 with both parents employed (or with an employed single parent)—47 percent are enrolled in preschool in California, compared to 60 percent elsewhere. The U.S.-California gap in preschool enrollment persists across every demographic group, including family structure, child’s ethnicity, mother’s education, and family income (Table 5.1). For virtually all of these demographic categories, California has both more children in groups less likely to participate in preschool and lower participation within the subgroup. For example, 26 percent of children ages 3 and 4 in California live in poverty, compared to 21 percent elsewhere. Throughout the United States, children from poor families are less likely to be enrolled in preschool, but the gap between U.S. and California preschool enrollments is larger for poor families than for any other income category. Fewer California children come from the high-income or high-education groups that were most likely to participate in preschool, but enrollment is lower in California even for these more advantaged groups. In looking at California only, we also find that regional differences as well as differences in the number of licensed child care slots in the county of residence are linked to differences in preschool enrollment. As Figure 5.3 shows, Los Angeles and the Farm Belt counties of the Central Valley and Central Coast have the lowest preschool enrollments, whereas the Bay Area and the North and Mountain counties have the highest enrollment.3 In part, this is linked to the licensed child care capacity in these regions, measured by the ratio of children to licensed child care slots (Figure 5.4).4 The Bay Area and the North and Mountain counties have the lowest ratio of children per licensed slot available, and Los ____________ 3See the appendix for a list of counties included in each region. 4This measure, developed by the California Child Care Resource and Referral Network, divides the children of working parents in the population (based on census counts) by the number of slots in centers and licensed family day care. The statewide average for this measure was 4.7 children (up to age 13) for each licensed slot. 44 Table 5.1 Percentage of Children Ages 3 and 4 in Preschool, by Demographic Group California Rest of United States Parental employment All parents employed Not all parents employed Parent works evenings/weekends Yes No Family structure Single parent Two parents Child’s race/ethnicitya Non-Hispanic white Hispanic Parental educationb Less than high school High school diploma/GED Some vocational education/college Associate degree or more Other adults in household No Yes Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare 47* 34 42* 49 52* 36 45 35* 30* 35 39* 55* 42 34* 29* 36* 48* 51 50* 33 40* 33* 60* 41 55* 62 57* 50 52 35* 31* 46 48* 64* 52 45* 42* 43* 52* 63 60* 40 47 53* NOTES: Italics indicate significance at the 95 percent level. All differences between California and rest of the United States were statistically significant. aSample sizes in California are too small to report breakouts for black and Asian children. bBased on years of education of the “most knowledgeable adult” for the focal child. *Indicates significant differences from reference group 45 Percentage 70 60 *57 50 *50 *43 40 *31 30 32 20 10 0 Bay Area Los Angeles Other Southern Farm Belt California North and Mountain * Indicates significance relative to the Farm Belt at the 95 percent level. Figure 5.3—Share of California Children Ages 3 and 4 in Preschool, by Region Percentage 70 60 Higher Capacity Lower *51 50 *50 40 30 *28 20 10 0 Fewer than 4 children per licensed slot 4–5 children per 5–6 children per licensed slot licensed slot 34 More than 6 children per licensed slot * Indicates significance relative to the lowest capacity group at the 95 percent level. Figure 5.4—Share of California Children Ages 3 and 4 in Preschool, by Child Care Capacity in County of Residence 46 Angeles has the highest ratio. Not surprisingly, the takeup of preschool is higher in counties with more child care slots available. However, we cannot distinguish between supply-driven differences (where enrollment is lower because capacity is lower) and demand-driven differences (where capacity is lower because fewer families are seeking preschool). The high availability of licensed child care in the North and Mountain counties is largely due to state-subsidized preschools in this region, along with relatively fewer other options. In this region, the high capacity reflects a large supply that may not be fully utilized. Because many of the characteristics in Table 5.1 occur together—for example, low-income families often have lower educational attainment— Table 5.2 provides regression results assessing the marginal contribution of different factors in explaining the likelihood that a child age 3 or 4 attends preschool in California.5 Some socioeconomic disadvantages reduce the chances of a child attending preschool. For example, children with less-educated mothers or from lower-income families are less likely to attend preschool. Income is one of the most important factors determining preschool attendance in California. Controlling for other factors, and for place of residence, the probability of attending preschool rose 4 percent for each additional $1,000 of family income. Compared to children whose mothers just completed high school or earned a GED, children whose mothers did not finish high school were 3 percent less likely to attend preschool, and those with more education were 3 to 4 percent more likely. Welfare recipients and low-income mothers who had never received welfare were 6 to 7 percent less likely than higher-income mothers to have their children enrolled in preschool. However, controlling for other factors, mothers who had previously received welfare were more likely than mothers with other welfare statuses to have their children in preschool— a result that is consistent with the observation, based on the use of child care vouchers, that families who receive child care assistance while on ____________ 5Consistent with Table 5.2, Appendix Table A.4 presents regression results for the United States with and without interaction terms for California. Appendix Table A.5 presents California regression results with and without regional variables. 47 Table 5.2 Regression Results on the Likelihood of California Children Ages 3 and 4 Being Enrolled in Preschool Household composition and employment Single parent, employed Married/partners, both employed full-time At least one parent not working full-time (married or single) Parent works evenings or weekends Family income For each additional $1,000 of annual family income Mother’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Mother’s welfare status Currently on welfare Formerly on welfare Never on welfare—income below 75% SMI Never on welfare—income above 75% SMI Other adults in home Hispanic child Child is age 3 Region of state Bay Area Los Angeles Other Southern California Farm Belt North and Mountain Children per licensed slot in the county % Change in Likelihood 36* 16* – 2* 4* – 3* 4* 3* – 6* 4* – 7* – 9* 5* – 16* 5* – 7* 8* – 4* – 5* NOTE: Italics indicate reference group. *Indicates significant differences at the 99 percent level. Marginal effects calculated from probit. welfare often shift to more structured care settings after they transition off welfare. Surprisingly, single parenthood does not appear to be a disadvantage in terms of preschool attendance. Compared to children of married parents who did not both work full-time, the children of single parents were 20 percent more likely to attend preschool. Obviously, married parents may choose to work part-time or not at all specifically to stay 48 home with young children. Yet even among unemployed single parents, the probability of their children attending preschool was dramatically higher. The explanation for this is not immediately clear. Without controlling for other factors, employed single parents are the most likely to have their children in preschool, followed by dual-income two-parent families. However, non-employed single parents are more likely than other non-working parents to have a child in both preschool or other care settings (Figure 5.5). Single parents may live in neighborhoods better served by preschools, perhaps because of greater population density in urban areas. Although single parents may also have greater access to free preschool through such programs as Head Start, low use of care by single parents in poverty suggests that such programs are not the primary drivers of this phenomenon. It is interesting to note that the much lower preschool participation for Hispanic children appears to be driven almost entirely by the economic and family characteristics of Hispanic families. Once we account for income, education, employment status, and other adults in the home, Hispanic children are actually more likely to attend preschool. Finally, the other effect that changes between the simple descriptive statistics and the regression findings is the preschool participation in the Percentage 80 70 65 69 52 50 46 40 38 30 21 20 Single parent, employed Single parent, not employed Two parents, both employed Two parents, one employed Two parents, not employed 73 44 30 30 23 17 31 18 10 0 Preschool 6 Other care 5 Not in care Figure 5.5—Child Care Settings for Children Ages 3 and 4, by Household Composition/Employment 49 North and Mountain counties. In Figure 5.3, we saw that this region had the second highest participation rate among the five California regions. However, once we control for other factors, especially the larger capacity of licensed slots in the region, the marginal role of this region of residence is negative; so all else being equal, children in this region are less likely to be enrolled in preschool. Targeting Efforts to Groups Less Likely to Use Preschool Efforts to improve preschool attendance in California may be most effective if they are tailored to address those populations least likely to use preschool. As we have seen, children ages 3–5 not in preschool fall into two categories: children in parental care only and children of working parents who are placed in other child care settings. In this section, we review the characteristics of each group to better understand the challenges in increasing preschool enrollment. Children in Parental Care Only Not surprisingly, the majority of children ages 3–5 who are not in any care setting are in two-parent families where one parent stays home, typically to serve as the primary caregiver. Preschool-age children are more likely to be cared for only by parents when there are other children in the household. Just 16 percent of only children are cared for exclusively by parents, compared to 28 percent of children with one sibling under age 13 and 38 percent of children with two or more siblings (Figure 5.6). This effect is particularly strong if the sibling is younger than age 3. At the same time, however, children cared for by parents only are more likely to come from disadvantaged households. Sixty-four percent of these children have mothers whose education ended at or before high school graduation, and a disproportionate share are Hispanic.6 Finally, these children are much more likely to be in low-income households. Poor children are 70 percent more likely to be cared for only by parents, ____________ 6Liang, Fuller, and Singer (2000) found that when controlling for economic and education attributes, the lower probability of Hispanic parents than whites using center care diminishes, as was the case with other adults in the household. 50 Percentage 80 70 60 57 50 50 40 38 30 20 10 0 Preschool One child under age 13 Two children Three or more children 27 22 24 38 28 16 Other care Not in care Figure 5.6—Child Care Settings for Children Ages 3–5, by Number of Children Under Age 13 largely at the expense of preschool attendance, and near-poor children are also less likely to be placed in a non-parental setting. In fact, for preschool attendees, the median household income is $41,700 whereas the child not placed in care comes from a household with a median income of $25,000. Assessing the characteristics of children not in child care arrangements through the prism of universal preschool, however, reveals another pattern that is important for school-readiness policies: Not only are these children less likely to attend preschool, but they are also less likely to attend kindergarten at age 5. Statewide, 54 percent of children age 5 attended kindergarten, compared to only 49 percent of low-income children (children from families below 50 percent of the state median income). A similar pattern holds for children of less educated mothers: 60 percent of children age 5 whose mothers had an associate degree or more education attended kindergarten, compared to only 49 percent of those whose mothers did not complete high school. Much of the difference is driven by differences in the Hispanic community. Only 43 percent of Hispanic children age 5 attended kindergarten, 16 percentage points lower than the share of non-Hispanics. (The lower use of kindergarten by Hispanic families holds outside California as well.) 51 Children in Other Care Settings Increasing preschool enrollment for children in other care settings is a somewhat different challenge. These families have chosen to place their children in a care setting, but whether because of the characteristics of the caregivers, the costs of care, or the convenience of the providers, they have selected settings other than preschool. In fact, 62 percent of these children are cared for by relatives, most commonly in the child’s home (Figure 5.7). Relatives may be the most affordable option for these families, but there may also be a strong preference for this type of care. Many of these relatives provide care in the child’s home. Working families are most likely to rely on care settings other than preschool. Use of other care settings is almost twice as common for children ages 3 and 4 with both parents working (or a working single parent) than for children with a parent at home: 38 percent versus 20 percent. Other characteristics associated with an increased likelihood of selecting care outside preschool settings include current welfare receipt, fewer children in the household, and lower income. An important factor in parents’ selection of settings other than preschool is the ability to have one consistent child care provider. More than half of children ages 3 and 4 in preschool regularly spend time in more than one care arrangement, compared to only 19 percent of children in other care settings. For families without a parent at home, 80 70 62 60 Percentage 50 40 36 30 20 12 10 0 Family day care Nanny/babysitter Care by relative Figure 5.7—Child Care Settings for Children Ages 3–5 Not in Preschool 52 the share of families who combine preschool with another setting rises from 52 percent to 58 percent. This may be because most preschool programs are not full-time, as shown in Table 5.3. Only 33 percent of children in preschool are in a program for more than 30 hours per week, and the majority are in programs for fewer than 20 hours per week. Half-time programs are particularly common for children with a parent at home, whereas full-time programs are more common for families with both parents in the workforce. Half the children in full-time preschool programs are still in other care settings, predominantly children without a parent at home. In light of the complexity of arrangements required when combining preschool with other care arrangements to accommodate work schedules, it is not surprising that many parents choose other care settings over preschool. Table 5.3 Share of Preschool Arrangements That Are Part-, Moderate-, and Full-Time Hours in Preschool Part-time (<20 hr/wk) Moderate time (20–29 hr/wk) Full-time (30+ hr/wk) % of Children Ages 3 and 4 in Preschool 52 15 33 % by Parents’ Employment Status Parent(s) Parent at Employed Home 43 64 16 15 48 20 NOTE: Children may be in more than one preschool arrangement. Promoting Use of Preschool California families are significantly less likely than families in the rest of the nation to enroll children ages 3–5 in preschool programs. Given that the Master Plan for California Education has identified rising concerns about school readiness for children entering kindergarten, our analysis of the characteristics of families of these children in California, their choice of care, and the costs of care points to some possibilities for increasing preschool enrollment in California. Although we do not have enough information to determine how reducing the costs of preschool would affect parents’ choice of settings, the findings in our study do 53 suggest several recommendations for encouraging enrollment in preschool: • Encourage kindergarten enrollment. Although California’s relatively early age cutoffs for kindergarten result in higher enrollment in California than elsewhere in the United States, kindergarten enrollment for children age 5 is disproportionately low among low-income families, children with less-educated parents, and especially among Hispanic families. • Focus on mothers with lower educational attainment. More than 83 percent of preschool-age children not attending preschool or kindergarten are cared for by their parents or other relatives. Some of these families may face economic barriers to enrolling their children in preschool, but controlling for income, mothers with lower educational attainment are less likely to place children in preschool. Targeted efforts to promote the use of preschool may be appropriate for some of these families, especially those who may not take advantage of opportunities offered by Head Start and other currently funded programs.7 • Offer more full-day preschool programs. State preschools in California traditionally offer half-day programs, which may not be attractive to working families. If this model is followed for universal preschool, families with both parents working full-time or poor single-parent families may continue to rely on care settings that better accommodate work schedules. Full-day fullyear programs or programs that help parents with multiple providers may be more attractive, although this strategy could be substantially more expensive. ____________ 7This recommendation assumes that programs such as Head Start and state contracted centers are available and accessible. 54 6. Costs of Universal Preschool The cost of care for universal preschool depends on a number of factors. First, there are the factors that determine how many children will enroll in preschool funded through a universal program, including eligibility rules and the share of eligible children who choose to enroll. A related issue is the share of children currently enrolled in privately paid child care who move into the newly funded program. Finally, there are the determinants of the costs per child. In the remainder of this chapter, we describe alternative assumptions that could be made for each of these factors and then explore the range of costs implied by the different scenarios. Assumptions on Takeup of and Costs for Preschool The first task in estimating the costs of universal preschool is to estimate the number of children likely to enroll. To determine this, we must first consider who would be eligible for universal preschool. Although “universal” suggests that all children would be eligible, the initial drive could target only low-income children. The Master Plan and proposed legislation suggest that children ages 3 and 4 would qualify. Presumably, children age 5 who do not meet the cutoff for kindergarten would also qualify. The second consideration is what share of parents will choose to enroll their eligible children, since many parents may continue to prefer their current care arrangements, whether parental care or in other informal settings. Of course, some of the children who would qualify for free care under the universal preschool plan are already enrolled in preschool. Takeup of the universal program by these children will increase program costs without increasing the number of children actually enrolled in preschool. How many will move into this program from existing preschool providers depends in part on which providers are included in the program. Some models of universal preschool rely on schools to offer the care; others permit existing child care centers to 55 qualify as providers. Typically, universal preschool programs have standards for providers that are more rigorous than licensing requirements. If providers can easily convert to a universal preschool program, then the children in their care would presumably shift from private payment and voucher assistance to the free program. If providers cannot easily convert, parents whose children are eligible for the universal program might still keep their children with these providers, because they prefer the provider (because of location, convenience, appropriateness for their child, or other reasons) or because they cannot access the free care. Finally, many preschool children already receive free care through Head Start, state preschools, state contracted slots, or subsidized care. We do not count these children as imposing new costs on the system, although they may shift between child care settings. We have considered different scenarios for each of these factors. In addition, we have considered different strategies for calculating the average costs per child in the program. Families Eligible for Universal Preschool We have examined four eligibility criteria for universal preschool. The broadest definition is “universal,” meaning all children of the appropriate age would be eligible to receive free preschool. At its broadest, this definition would include all children ages 3–5 in California, except those age 5 enrolled in kindergarten. A more constrained version would limit eligibility to those families currently eligible for child care subsidies—a definition that would include only children from families with income below 75 percent of the state median income. However, because universal preschool is focused on child development, children would be eligible whether or not their parents are in the workforce, a difference from the subsidy system. Our third option is a variation on this, limiting eligibility to children with family income below 50 percent of the SMI. Current subsidy policy does not require co-payments for families in this income range. Finally, we consider narrowing eligibility to children with family income below poverty. These children are already eligible for preschool through the Head Start program but slots are not necessarily available to serve all children. 56 Enrollment of Children Not Currently in Preschool One would never expect all eligible children to actually enroll in a voluntary universal preschool program. Short of 100 percent enrollment, we have developed two alternative scenarios for the takeup of universal preschool. Our high estimate assumes that preschool enrollment for participating children matches that of wealthy families. To make this truly an upper bound, we have chosen a higher income threshold than the median income, which defined our top income category above. For this exercise, we have selected an income threshold of 150 percent of the SMI to define a group of children whose parents faced much lower economic barriers to preschool enrollment. We are assuming, essentially, that in the absence of economic barriers to preschool enrollment, the underlying preferences for preschool over other care options are identical between high- and low-income families.1 Our low estimate was set arbitrarily at half of our high estimate. To calculate enrollment based on the high estimate assumption, we have compared enrollment of California families qualifying under each eligibility scenario to enrollment for children of the same age in families with income above 150 percent of the state median income. Crowd-Out of Existing Preschool If universal preschool were available, many if not most children currently enrolled in preschool would switch to the universal preschool program. For our high estimate, we assume that all children enrolled in preschool paid for by their parents would move from privately paid preschool to the new program. Children already enrolled in free preschool programs, such as Head Start, are not included in our count of new enrollment costs. Existing free preschool care may become part of the universal program, but this would not incur new costs. Therefore, our high estimate uses the NSAF data to determine the share of families (in different eligibility groups) enrolled in preschool care with any outof-pocket child care expenses. This is probably an upper bound for two ____________ 1Realistically, the cost of preschool is still perceived as a potential barrier to enrollment for higher-income families. Nor do we have a good understanding of what different families’ preferences would be in the absence of any cost for preschool. 57 reasons. First, because of small sample sizes in the various eligibility groups, we use U.S. estimates of the share of children in paid preschool care, which is higher than the share paying for care on average in California. Second, because out-of-pocket costs are recorded for all care settings and all children, it is possible that we are overcounting the share of families paying for preschool when their children are in multiple settings. Some private paying parents may in fact choose to keep paying for preschool even when their children are eligible for the new program. How many will depend on the way the universal preschool is implemented. For example, many more would switch if most existing preschool providers are able to participate in the universal program. Fewer would switch if the universal care is only half-day or provided only in limited locations. We assume that higher-income families are less likely to switch than lower-income families. Thus, our low estimate starts with the share paying for preschool in California. We then assume that 50 percent of families with income above 75 percent of the SMI will switch to the new program, as will 70 percent of families between the poverty threshold and 75 percent of SMI and 90 percent of poor families. Using these assumptions, Table 6.1 shows a variety of scenarios for takeup of universal preschool by eligibility criteria and age. It also shows our two assumptions for the share of children currently in preschool who would also move to free care. Because poorer families are the least likely to participate in child care currently, a larger share of children would need to move from parental or other care into preschool to match the enrollment of wealthier families. The takeup estimates are fairly similar for the two low-income eligibility standards—around 60 percent. For all families to match the preschool use of wealthy families, 53 percent of children not currently in preschool would have to move into such arrangements. The higher share for children age 5 reflects the lower use of both preschool and kindergarten by lower-income families. Some of these children would be eligible for kindergarten. However, we assume that a somewhat larger share enter preschool instead, assuming that parents of these children feel that their children are not ready for school. 58 Table 6.1 Assumptions on Takeup of Universal Preschool for California Children All families eligible All preschoolers Age 3 Age 4 Age 5 Eligible if income below 75% SMI All preschoolers Age 3 Age 4 Age 5 Eligible if income below poverty All preschoolers Age 3 Age 4 Age 5 aExcludes kindergartners. Enrollment as % of Children in Care Other Than Preschoola High Low Estimate Estimate 53 27 51 25 53 27 62 31 58 29 57 29 57 28 68 34 60 30 59 29 57 29 68 34 % of Children in Preschool Moving to Free Care High Low Estimate Estimate 62 37 50 26 38 25 Among families already using preschool, low-income families are much more likely to have free care. For this reason, the more constrained the eligibility standards, the smaller the share of children moving into the new program. Thus, more generous eligibility standards increase not only the total number of children who could switch from family-paid preschool to state-funded preschool but also the share of those we would expect to switch. Cost per Child in Universal Preschool Policymakers would have more control over the cost per child for universal preschool, although they would still face constraints. In particular, the cost per child would have to be sufficient to encourage the supply expansion required to absorb all the additional children served. In addition, the costs would need to cover care of a sufficiently high quality to justify choosing preschool over other care arrangements. There are three potential models for the cost of care: market prices, state 59 preschool allocations, or the standard reimbursement rate (SRR) provided for children in contracted slots in child care centers. State preschool is a part-day, part-year program, whereas the contracted slots are usually full-day, full-year.2 Similarly, parents in the broader child care market use both full-time and part-time care. Alternative Estimates of the Costs of Universal Preschool Using the set of options described above, Tables 6.2 and 6.3 show the number of children served and the costs of care in our various scenarios based on the number of children ages 3–5 in the 2000 Census. We consider five options for the price of care per child. The first three are market-based prices. The NSAF estimate of $4,216 annually Table 6.2 Estimates of Average Annual Costs for Universal Preschool: High Estimate ($ millions) All families eligible All preschoolers Age 3 Age 4 Age 5 Eligible if income below 75% SMI All preschoolers Age 3 Age 4 Age 5 Eligible if income below poverty All preschoolers Age 3 Age 4 Age 5 No. of Children Served 726,402 272,626 301,308 152,468 404,493 169,524 161,701 73,268 205,163 76,790 87,211 41,162 NSAF $4,216/ yr 3,062 1,149 1,270 643 1,705 715 682 309 865 324 368 174 RMR Survey Part-Time Full-Time $3,200/yr $5,800/yr 2,324 872 964 488 4,213 1,581 1,748 884 1,294 542 517 234 2,346 983 938 425 657 1,190 246 445 279 506 132 239 State Preschool 2,348 881 974 493 1,308 548 523 237 663 248 282 133 SRR 4,881 1,832 2,025 1,025 2,718 1,139 1,087 492 1,379 516 586 277 ____________ 2For simplicity, we do not adjust assumptions on the likelihood of switching from privately paid preschool to the universal program to reflect whether the proposed program is part-day, part-year or full-day, full-year. 60 Table 6.3 Estimates of Average Annual Costs for Universal Preschool: Low Estimate ($ millions) All families eligible All preschoolers Age 3 Age 4 Age 5 Eligible if income below 75% SMI All preschoolers Age 3 Age 4 Age 5 Eligible if income below poverty All preschoolers Age 3 Age 4 Age 5 No. of Children Served 401,335 145,673 167,649 88,013 206,242 85,567 82,930 37,745 110,825 39,621 48,316 22,888 NSAF $4,216/ yr 1,692 614 707 371 869 361 350 159 467 167 204 96 RMR Survey Part-Time Full-Time $3,200/yr $5,800/yr 1,284 466 536 282 2,328 845 972 510 660 1,196 274 496 265 481 121 219 355 643 127 230 155 280 73 132 State Preschool 1,298 471 542 284 667 277 268 122 358 128 156 74 SRR 2,697 979 1,127 591 1,386 575 557 254 745 266 325 154 for a child in preschool averages across children in part-time and fulltime settings.3 Two other estimates of the market price come from the regional market rate (RMR) survey conducted annually through 2001 to determine price ceilings for subsidy rates. The RMR survey collects data on child care fees for centers in every county in California, taking a census in small counties and a sample in larger counties. The survey reports fees separately for full-time and part-time care. These two options are shown as different scenarios for the price estimates. The part-time preschool rate from the RMR survey is nearly identical to the part-day, part-year rate provided through state preschools. The final cost option assumes the full-year, full-day SRR for state contracted slots. ____________ 3Because the NSAF reports costs per family across all arrangements, we use the U.S. estimates for the cost of care for children without siblings in preschool as the only setting. (Taking children in one setting only biases this price toward part-day programs.) The U.S. estimate is then inflated to account for average price differences between California and the rest of the United States. 61 The exact method for selecting the cost of care matters less than the distinction between full-time and part-time care: The estimates based on the SRR are very close to those based on the full-time RMR, and the estimates based on state preschool funding are very close to those based on the part-time RMR. Because the NSAF captures a blend of full-time and part-time care, the estimates based on the NSAF fall between the other pairs. A universal preschool program serving all preschool children fulltime, full-year could be expected to cost as much as $4.88 billion annually, assuming high takeup rates (Table 6.2). The lower estimates of takeup drop this to $2.70 billion (Table 6.3). As the eligibility criteria narrow—moving the program away from universal coverage and closer to a poverty program—the costs fall. A program supporting full-year, fulltime preschool care for children in poverty would cost between $1.38 billion (Table 6.2) and $745 million (Table 6.3). Part-day, part-year programs such as the current state preschools would cost about half as much as the full-time programs, with the costs at the broadest eligibility falling to between $2.35 billion and $1.3 billion. However, our estimates do not adjust for lower takeup of parttime care, given the reduced convenience for working families. Programs that permitted parents to choose between full-time and part-time care would fall between these cost estimates and would also have the highest expected enrollment rates. A substantial portion of these costs would go toward serving families who are already enrolling their children in preschool. For the three eligibility options, Table 6.4 shows the share of costs resulting from children currently enrolled in preschool. The share of costs from currently enrolled children falls as eligibility is constrained, both because poorer children are more likely to receive free preschool in the absence of a universal preschool program and because total enrollment in preschool rises with income. Below the 75 percent of SMI threshold, the different assumptions on takeup by children already enrolled in preschool have relatively little effect. If all families are eligible, this share ranges from 50 to 54 percent; for children in families with income below 75 percent of SMI, the share of costs from currently enrolled preschoolers falls to about 34 percent. In fact, the high and low estimate assumptions matter 62 Table 6.4 Percentage of Costs Resulting from Participation of Currently Enrolled Preschoolers, by Takeup Assumptions Eligibility Criteria All families eligible Income below 75% SMI Income below poverty High Estimate 50 33 26 Low Estimate 54 34 32 somewhat more for the takeup by children not currently enrolled in preschool, so the share of costs resulting from currently enrolled children is slightly higher in the low estimate version. Finally, it is important to recognize that these cost assumptions abstract from the issue of available supply. In fact, it may be necessary to increase reimbursement rates to induce a sufficient increase in supply of preschools to accommodate children served under this program. Table 6.5 reports the additional slots that would be needed in our high and low takeup scenarios and for each eligibility criterion. These counts include children currently in parental care only or in settings other than preschool and kindergarten. Obviously, the number of additional slots needed is higher the broader the eligibility criteria and the higher the takeup by children not currently enrolled. (In our method, these factors are not affected by the choice of reimbursement rate.) The extra slots required range from around 76,000 in the lowest estimates to 366,000 in the highest estimates. This massive addition of slots could not be accommodated quickly by existing providers. Table 6.5 Additional Slots Needed for Children Enrolling in Preschool, by Takeup Assumptions Eligibility Criteria All families eligible Income below 75% SMI Income below poverty High Estimate 365,895 270,774 151,606 Low Estimate 182,947 135,387 75,803 63 Summary The costs of universal preschool vary widely depending on assumptions of eligibility criteria and takeup of preschool by children not currently enrolled. At the high end, full-year, full-day universal preschool could cost as much as $5 billion annually when fully implemented, although eligibility criteria more closely matched to the current subsidy system would reduce this cost to between $1.4 billion and $2.7 billion. Of course, the broader the criteria, the less the program may be perceived as a public assistance program and presumably the higher its political popularity. Similarly, if universal preschool were modeled after state preschools, offering part-day, part-year programs, it would be significantly less expensive. The method for estimating the average cost of care matters less than the distinction between full-time and part-time programs. Parttime programs are much less accessible to working parents. Given the significant gap in preschool use between high-income and low-income families with a parent not in the workforce, this limitation may be a secondary concern. Finally, it is critical to note that these costs do not account for the administrative and other infrastructure costs that would be required to run a universal preschool program. In particular, if large numbers of children do move into preschool, there may be substantial startup costs to ensure an adequate number of facilities and classrooms to accommodate the new enrollment. 64 7. Conclusions As state resources become ever more limited, it is difficult to maintain the existing balance between child care policy intended to support working parents and policy to promote developmentally enriched care. At present, child care subsidies provide an important resource to parents who need child care to permit them to work. These subsidies have been a particularly important resource for families transitioning from welfare to work. These families, as with other lowincome families, spend a disproportionate share of their income on child care. Yet many families who are eligible for child care do not receive government assistance because the pool of subsidy funds is not sufficient to cover all eligible children. Without additional funding, the only way to support more children is to provide less support to those currently covered. Since assistance seems to encourage families to use more formal, presumably more enriched care, and such options as increased family fees would provide disincentives for families choosing higher-priced care, this tradeoff could also discourage assisted families from using high-quality care. At the same time, such proposals as the call for universal preschool are seeking to substantially increase spending on developmentally rich early childhood education programs. Although the existing subsidy system offers substantial parental choice, universal preschool would provide very generous benefits for enrolling in high-quality programs. The hope is that increased school readiness preparation, especially for disadvantaged children, would pay off down the road in improved school outcomes. Advocates of universal preschool view the link to school readiness and universal coverage as important selling points for these proposals. Nevertheless, universal preschool may cost several billion dollars annually, with as much as 50 percent of the cost offset by private dollars spent by families with children currently enrolled in preschool. 65 The tradeoff between these two goals of work support and developmental care is a political choice. Whatever is chosen, it is important to address the needs of disadvantaged families, which are paying 20 percent or more of their earnings toward child care with relatively low access to the more formal care sectors. In addition, the low participation of children of parents with low educational attainment and of Hispanic children in preschool or other care settings suggests that to improve school readiness for these populations, additional efforts may be needed to increase awareness of the advantages of formal programs of early childhood education. 66 Appendix Supplementary Tables Table A.1 Sample Characteristics for Family-Level Data (Costs) Most knowledgeable adult’s race/ ethnicity Non-Hispanic white Hispanic Other Income as a % of federal poverty threshold <100 100–199 200+ Welfare history Currently on welfare Formerly on welfare Never on welfare Most knowledgeable adult’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Number of children under age 13 1 2 or more California Weighted No. % Rest of United States Weighted No. % 483 41 14,004 69 599 40 2,856 11 201 19 3,738 20 364 26 4,394 20 344 23 5,335 23 575 51 10,599 57 154 11 1,220 6 156 10 3,048 13 966 79 15,926 81 264 21 2,211 11 357 25 6,351 32 292 23 4,705 23 348 31 6,842 33 402 40 7,011 44 881 60 13,279 56 67 Table A.2 Percentage of Children Ages 0–5 in Rest of United States in Child Care, by Demographic Group and Type of Arrangement % in NonParental Care % in Structured Care Care by Relative in Child’s Home % in Arrangementa Care by Relative Outside Child’s Family Home Day Care Nanny/ Babysitter 68 Family structure Single parent Two parents Child’s race/ethnicityb Non-Hispanic white Hispanic Parental educationc Less than high school High school diploma/GED Some vocational education/college Associate degree or more Other adults in household No Yes Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI 81 70 74 63 57 71 74 79 74 73 64 69 74 81 58 57 57 48 49 53 56 63 59 48 56 51 57 62 32 23 24 31 32 28 26 21 21 50 30 28 24 21 29 20 7 26 21 14 27 22 15 30 14 9 30 11 5 31 21 7 28 23 10 22 21 19 28 21 13 21 16 4 30 15 8 31 20 8 27 22 10 22 23 17 Table A.2 (continued) 69 Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % in NonParental Care 79 65 75 62 % in Structured Care 60 49 61 65 Care by Relative in Child’s Home 22 29 30 28 aChildren may be in more than one arrangement. bSample sizes in California are too small to report breakouts for black and Asian children. cBased on years of education of the “most knowledgeable adult” for the focal child. % in Arrangementa Care by Relative Outside Child’s Family Home Day Care 23 22 31 19 30 18 26 13 Nanny/ Babysitter 16 9 7 6 Table A.3 Counties Included in Each Region Other Southern Bay Area Los Angeles California Farm Belt Alameda Los Angeles Orange Colusa Stanislaus Contra Costa Riverside El Dorado Sutter Marin San Bernardino Fresno Tulare Napa San Diego Glenn Yolo San Francisco Santa Barbara Imperial Yuba San Mateo Ventura Kern Santa Clara Kings Santa Cruz Madera Solano Merced Sonoma Monterey Placer Sacramento San Benito San Joaquin San Luis Obispo North and Mountain Alpine Plumas Amador Shasta Butte Sierra Calaveras Siskiyou Del Norte Tehama Humboldt Trinity Inyo Tuolumne Lake Lassen Mariposa Mendocino Modoc Mono Nevada 70 Table A.4 Regression Results on Likelihood of Preschool Children Ages 3 and 4 Being Enrolled in California and Rest of United States Household composition and employment Single, employed Married/partners, both employed full-time At least one parent not working fulltime (married or single) Parent works evenings or weekends Income is $1,000 higher Mother’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Mother’s welfare status Currently on welfare Formerly on welfare Never on welfare—income below 75% SMI Never on welfare—income above 75% SMI Other adults in home Hispanic child Child is age 3 California resident % Change in Likelihood of Child Being in Preschool Rest of United Rest of States (No United States California Interaction) Interactions Interactions 12* 27* 18* – 1* 11* 7* – 5* 1* – 8* 3* 12* 3* – 0.5* – 11* – 6* 1* – 8* 3* 13* 3* – 2* – 12* 4* 3* 6* 1* – 8* – 8* 6* 5* – 9* – 9* – 22* – 10* – 9* – 14* – 23* – 49* 16* 2* 8* –– NOTE: Italics indicate reference group. *Indicates statistically significant at the 99 percent level. Marginal effects are calculated from probit. 71 Table A.5 Regression Results on Likelihood of Children Ages 3 and 4 Being Enrolled in Preschool in California Household composition and employment Single, employed Married/partners, both employed full-time At least one parent not working full-time (married or single) Parent works evenings or weekends Income is $1,000 higher Mother’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Mother’s welfare status Currently on welfare Previously on welfare Never on welfare—income below 75% SMI Never on welfare—income above 75% SMI Other adults in home Hispanic child Child is 3 Region of state Bay Area Los Angeles Other Southern California Farm Belt North and Mountain Children per licensed slot in the county % Change in Likelihood of Child Being in Preschool No With Regional With Regions Variables Regions and Slots 40* 38* 36* 20* 18* 16* – 1* 4* – 2* 3* 6* – 4* 4* – 7* – 13* 4* – 15* – 2* 4* – 4* 4* 3* – 7* 4* – 7* – 9* 6* – 15* 12* – 11* 4* 0.5* – 2* 4* – 3* 4* 3* – 6* 4* – 7* – 9* 5* – 16* 5* – 7* 8* – 4* – 5* NOTE: Italics indicate reference group. *Indicates statistically significant at the 99 percent level. Marginal effects are calculated from probit. 72 Bibliography Brick, J. Michael, Pam Broene, David Ferraro, Tom Hankins, and Teresa Strickler, 1999 NSAF Sample Estimation Survey Weights, Urban Institute, Washington, D.C., July 2000, available at www.urban.org/ UploadedPDF/1999_Methodology_3.pdf. Brick, J. Michael, Gary Shapiro, Ismael Flores-Cervantes, David Ferraro, and Teresa Strickler, 1997 NSAF Snapshot Survey Weights, Urban Institute, Washington, D.C., July 2000, available at www.urban.org/ UploadedPDF/Methodology_3.pdf. California State Senate, Joint Committee to Develop a Master Plan for Education, The California Master Plan for Education, 2002, available at www.sen.ca.gov/masterplan. Cappiziano, Jeffrey, Sarah Adelman, and Matthew Stagner, What Happens When the School Year Is Over? The Use and Cost of Child Care for School-Age Children During the Summer Months, Urban Institute, Washington, D.C., June 2002, available at www.urban.org/ UploadedPDF/310497_OP58.pdf. Giannarelli, Linda, Sarah Adelman, and Stephanie Schmidt, Getting Help with Child Care Expenses, Urban Institute, Washington, D.C., February 2003, available at www.urban.org/UploadedPDF/310615_ OP62.pdf. Gomby, Deanna S., Mary B. Larner, Carol S. Stevenson, Eugene M. Lewit, and Richard E. Behrman, “Long-Term Outcomes of Early Childhood Programs: Analysis and Recommendations,” The Future of Children, Vol. 5, No. 3, Winter 1995. Hirshberg, Diane, Danny Shih-Cheng Huang, and Bruce Fuller, Welfare to Work and Child Care Selection: Which Families Use Subsidies and Home-Based or Center Care? Policy Analysis for California Education, 73 University of California, Berkeley and Stanford University, Berkeley, California, November 2002. Karoly, Lynn A., Peter W. Greenwood, Susan S. Everingham, Jill Houbé, M. Rebecca Kilburn, C. Peter Rydell, Matthew Sanders, and James Chiesa, Investing in Our Children: What We Know and Don’t Know About the Costs and Benefits of Early Childhood Interventions, RAND, Santa Monica, California, 1998, available at www.rand.org/ publications/MR/MR898/. Liang, Xiaoyan, Bruce Fuller, and Judith D. Singer, “Ethnic Differences in Child Care Selection: The Influence of Family Structure, Parental Practices, and Home Language,” Early Child Research Quarterly, Vol. 15, No. 3, 2000, pp. 357–384, available at www-gse.berkeley.edu/ research/PACE/ethnic_differences.pdf. Marrufo, Grecia, Margaret O’Brien-Strain, and Helen Oliver, Child Care Price Dynamics in California, Public Policy Institute of California, San Francisco, California, 2003. Sonenstein, Freya, Gary Gates, Stephanie Schmidt, and Natalya Bolshun, Primary Child Care Arrangement of Employed Parents, Urban Institute, Washington, D.C., May 2002, available at www.urban.org/ UploadedPDF/310487_OP59.pdf. 74 About the Authors MARGARET O’BRIEN-STRAIN Margaret O’Brien-Strain is an adjunct fellow at PPIC and Director of Social Policy at The SPHERE Institute. She is interested in a broad range of social services issues and has written about child care, home visitation programs, performance measurement, and welfare reform in California. She holds a B.A. in economics from Swarthmore College and a Ph.D. in economics from Stanford. LAURA MOYE Laura Moyé, formerly a research associate at PPIC, is interested in social issues related to child and youth well-being. She currently works in local government as a research analyst on child and youth development policy and programming. She holds a B.A. in urban studies from San Francisco State University and is currently pursuing an M.P.A. with a focus in policy analysis. FREYA LUND SONENSTEIN Freya Lund Sonenstein is Director of the Center for Adolescent Health at the Johns Hopkins University Bloomberg School of Public Health. She formerly directed the Population Studies Center at the Urban Institute in Washington, D.C., and was a visiting fellow at the Public Policy Institute of California in 2001. She is a Wellesley College graduate and holds a Ph.D. in sociology from Boston University. 75 Related PPIC Publications The Well-Being of California’s Children Frank F. Furstenberg, Maureen R. Waller, and Hongyu Wang What Happens to Families When They Leave Welfare? Thomas MaCurdy, Grecia Marrufo, and Margaret O’Brien-Strain Child Care Price Dynamics in California Grecia Marrufo, Margaret O’Brien-Strain, and Helen Oliver Expensive Children in Poor Families: The Intersection of Childhood Disabilities and Welfare Marcia K. Meyers, Henry E. Brady, and Eva Y. Seto “California’s Young Children: Demographic, Social and Economic Conditions” California Counts: Population Trends and Profiles Volume 4, Number 2, November 2002 Deborah Reed and Amanda Bailey A Portrait of Race and Ethnicity in California: An Assessment of Social and Economic Well-Being Belinda I. Reyes (editor), Jennifer Cheng, Elliot Currie, Daniel Frakes, Hans P. Johnson, Elizabeth Bronwen Macro, Deborah Reed, Belinda I. Reyes, José Signoret, and Joanne Spetz (contributors) Child Support and Low-Income Families: Perceptions, Practices, and Policy Maureen Waller and Robert Plotnick PPIC publications may be ordered by phone or from our website (800) 232-5343 [mainland U.S.] (415) 291-4400 [Canada, Hawaii, overseas] www.ppic.org 77" } ["___content":protected]=> string(104) "

R 1203MOR

" ["_permalink":protected]=> string(79) "https://www.ppic.org/publication/arranging-and-paying-for-child-care/r_1203mor/" ["_next":protected]=> array(0) { } ["_prev":protected]=> array(0) { } ["_css_class":protected]=> NULL ["id"]=> int(8278) ["ID"]=> int(8278) ["post_author"]=> string(1) "1" ["post_content"]=> string(0) "" ["post_date"]=> string(19) "2017-05-20 02:36:11" ["post_excerpt"]=> string(0) "" ["post_parent"]=> int(3429) ["post_status"]=> string(7) "inherit" ["post_title"]=> string(9) "R 1203MOR" ["post_type"]=> string(10) "attachment" ["slug"]=> string(9) "r_1203mor" ["__type":protected]=> NULL ["_wp_attached_file"]=> string(13) "R_1203MOR.pdf" ["wpmf_size"]=> string(6) "643748" ["wpmf_filetype"]=> string(3) "pdf" ["wpmf_order"]=> string(1) "0" ["searchwp_content"]=> string(141401) "Arranging and Paying for Child Care ••• Margaret O’Brien-Strain Laura Moyé Freya Lund Sonenstein 2003 PUBLIC POLICY INSTITUTE OF CALIFORNIA Library of Congress Cataloging-in-Publication Data O’Brien-Strain, Margaret. Arranging and paying for child care / Margaret O’Brien-Strain, Laura Moyé, Freya Lund Sonenstein. p. cm. Includes bibliographical references. ISBN: 1-58213-047-7 1. Child care services—California. 2. Preschool children— Care—California. I. Moyé, Laura. II. Sonenstein, Freya L. III. Title. HQ778.65.C2O27 2003 362.71'2'09794—dc22 2003023308 Copyright © 2003 by Public Policy Institute of California All rights reserved San Francisco, CA Short sections of text, not to exceed three paragraphs, may be quoted without written permission provided that full attribution is given to the source and the above copyright notice is included. PPIC does not take or support positions on any ballot measure or on any local, state, or federal legislation, nor does it endorse, support, or oppose any political parties or candidates for public office. Research publications reflect the views of the authors and do not necessarily reflect the views of the staff, officers, or Board of Directors of the Public Policy Institute of California. Foreword This report by Margaret O’Brien Strain, Laura Moyé, and Freya Lund Sonenstein provides a useful and timely analysis of child care for young children in California. What share of families use child care, and what type of care? How much public assistance do they receive to secure the care? How does California’s experience compare with that in the rest of the nation? How expensive would a universal preschool program be, and what are the alternatives to such a program? These are all questions central to any state governor facing growing budget deficits, and especially to a governor with an avowed interest in the future of the state as manifested in its children. The authors conclude that California families are less likely than families elsewhere in the United States to use child care of any kind for their young children. This is especially true for low-income families and those with less-educated parents. Children age 4 in California are also much less likely than those in the rest of the nation to attend preschool and, not surprisingly, this is especially true for children in lower-income families. Finally, the authors estimate that a full-year, full-day universal preschool care program would cost as much as $5 billion annually. Given that such a program is unlikely to emerge from the current budget crisis, or anything approaching a major increase in child care expenditures, the authors suggest that other options are likely to be on the table—for example, encouraging kindergarten enrollment for lowincome children, supported by Head Start. All options will be considered when welfare reform legislation is reauthorized in Washington in the next year or two. Encouraging parents to leave welfare is one thing. Finding the ways and means to support child iii care assistance to help working families is yet another. It is obvious that the first strategy cannot be successful without some substantial help from the other. David W. Lyon President and CEO Public Policy Institute of California iv Summary California’s child care policies represent a complex set of programs attempting to address the goals of both supporting working parents and promoting child development. As California faces a budget crisis and a major political change, these policies may be subject to a variety of changes. The 2003–2004 state budget included some reductions to the system of child care subsidies, which at its peak provided over $1.5 billion to help working families pay for child care, especially families transitioning from welfare to work. Governor Davis proposed much more drastic changes to the subsidy programs, but the most serious review of these policies is expected to follow the delayed federal reauthorization of welfare reform. Current federal proposals call for much more stringent work requirements, and the level of federal support for child care has been one of the stumbling blocks for passage of reauthorization legislation. At the same time, California has a number of programs aimed more toward promoting early childhood education, including state-funded preschools, contracted slots in child development programs, services to increase the quality of care, and resource and referral services. Although programs such as Head Start are built on the recognition that children at risk for school failure could benefit greatly from early childhood programs, recent pressures to improve school outcomes have turned public attention to the role of early childhood programs in preparing all children for school. Unlike Head Start or state-subsidized preschools that target groups based on income, universal preschool initiatives seek to promote access to quality early childhood programs for all preschool-age children on a voluntary basis. Thus, the California Master Plan for Education calls for universal preschool for children ages 3 and 4, a concept embraced by the David and Lucile Packard Foundation, First 5 California, and several county First 5 commissions, including Los v Angeles and San Mateo, which are all supporting exploratory efforts to implement such a program. In light of potential changes in child care policy, this report examines the current state of child care for young children in California. In particular, we address four key questions: • What share of California families use what types of child care for their young children, and how does this differ from the rest of the United States? • How much of their household income do families pay for child care and how much assistance do they receive in paying for it? • Are there major gaps between California and the rest of the country in preschool enrollment? If so, how much is explained by demographic differences? • If preschool access is expanded through universal preschool programs, how expensive will these programs be and how do different designs affect the cost of care through such programs? The analysis combines data on families with children ages 0–5 in the 1997 and 1999 rounds of the National Survey of America’s Families, for a total of 1,287 children ages 0–5 from 1,282 California families. We consider a broad range of care settings regularly used by parents of young children: structured care in programs such as child care centers, nursery schools, preschools, and Head Start programs; family day care provided to a group of children in the provider’s home; care by relatives in the child’s home or the relative’s home; and nanny or babysitter arrangements where an unrelated caregiver regularly provides child care in the child’s home. We examine all regular arrangements used by both working and non-working parents. Our key findings are listed below. • California families are less likely to use child care for their young children than are families elsewhere in the United States. In part, this is due to lower labor force participation by California parents and a larger share of families in groups that are lower users of child care. Even within these groups, however, child care use is lower in California than in the rest of the United States. Still, most children of working parents (83 percent) vi regularly spend time in non-parental care, averaging 35 hours per week. Forty-five percent of children with a parent not in the labor force are also regularly placed in non-parental care, albeit for fewer hours per week than the children of working parents. • Children with less-educated parents and poorer children are less likely to be placed in care. Because their families are disproportionately poor and less educated, Hispanic children in particular are less likely to be placed in care. Higher-income families are more likely than lower-income families to place children in structured care settings, although families formerly on welfare have patterns of child care use very similar to those of higher-income families. • Children of working single parents spend the most time in care and are also the most likely to be in multiple care arrangements. Although multiple care arrangements are common even for children with parents out of the labor force, the ability to coordinate work schedules has a strong influence on the number of arrangements used for young children, as well as on their hours in care. Families with some non-standard work hours can limit both the number of hours in care and the number of arrangements. Such families have greater reliance on care provided by relatives when they do use child care. • Families who pay for child care while parents are working pay an average of $373 per month out-of-pocket, representing about 10 percent of family earnings. This cost is higher than the average in the rest of the United States but approximately the same share of income. Poor families who do pay for care pay an average of 24 percent of earnings, compared to about 7 percent for families with income above the state median. About 28 percent of all working parents pay no out-of-pocket costs for child care. • Most families at or below 75 percent of the state median income— the eligibility cutoff for child care subsidies, about $38,000—receive some assistance in paying for child care, although often as free care provided by relatives. Forty-six percent of poor families, onethird of former welfare recipients, and two-thirds of current recipients receive government or social services assistance in vii paying for care. However, current welfare recipients are less likely to receive free care from relatives, perhaps because the welfare system will reimburse relatives for providing care. Use of structured care is much higher among welfare recipients and other low-income families receiving assistance. • Children age 4 in California are much less likely than those in the rest of the nation to attend preschool. Only 50 percent of these children in California are enrolled in preschool, compared to 63 percent in the rest of the United States. A similar gap holds for children age 3, but California’s relatively early kindergarten enrollment age makes up for the U.S.-California gap in preschool attendance for children age 5. • As with child care use overall, children with less-educated mothers or from lower-income families are less likely to attend preschool. Families with more familial support, from stay-at-home mothers, other adults in the home, or other available relatives, are also less likely to use child care. Together these factors help create a particular gap in preschool enrollment of Hispanic children: only 35 percent of Hispanic children ages 3 and 4 attend preschool compared to 45 percent of non-Hispanic children. • Full-year, full-day universal preschool care could cost as much as $5 billion annually when fully implemented, although if eligibility were limited to families with income below the cutoff for the current subsidy system, this cost would fall to between $1.4 billion and $2.7 billion. The costs of universal preschool vary widely depending on assumptions of eligibility criteria and on takeup of preschool by children not currently enrolled, but a substantial share of the costs would likely come from the shift of children currently enrolled in preschool from private payment to public funding. Part-day preschool would be substantially less expensive, around $1.3 billion to $2.3 billion for a universal program. Although part-time programs are much less accessible to working parents, the lower use of preschool by low-income families with a parent not in the workforce may make this limitation a secondary concern. Costs of implementing and administering the program are not included in these estimates. viii The findings in this report point to other options, short of universal preschool, for improving early childhood education. One important first step would be encouraging kindergarten enrollment. The same families least likely to use preschool are also disproportionately likely to have children age 5 not enrolled in kindergarten. This is particularly an issue for low-income and Hispanic families. Many children who would be eligible for programs such as Head Start are being cared for by their mothers or other relatives; outreach to and education of these groups could increase their involvement in the existing system. However, even with outreach, access to Head Start and state preschool programs would have to be addressed to fully accommodate eligible families. Finally, some families appear to select settings other than preschool because partday programs, such as those offered by the state preschools, do not match well with the needs of working parents. Nor should calls for universal preschool draw attention away from the ongoing need for child care assistance to help working parents. Reducing eligibility for subsidies will increase the costs of work for many families, increasing the share of income spent on child care. Although relatives caring for young children for free may absorb some of the costs, reduced subsidies may also force some parents currently using structured care settings to switch to more informal arrangements. If welfare reauthorization further increases the work requirements for welfare recipients, additional resources will be needed to assist families. ix Contents Foreword ..................................... Summary..................................... Figures ...................................... Tables ....................................... Acknowledgments ............................... iii v xiii xv xvii 1. INTRODUCTION ........................... 2. DATA .................................... The National Survey of America’s Families ............ Sample of Families with Young Children .............. Analysis Variables............................. Child Care Arrangements ...................... Child Care Expenses and Assistance ............... Economic and Demographic Variables .............. 3. OVERVIEW OF CHILD CARE USE IN CALIFORNIA .. Share of Children Under Age 6 in Child Care ........... Choice of Arrangements ........................ Choice of Arrangements, by Child and Family Characteristics ........................... Work Schedules and Care Arrangements .............. Summary .................................. 4. PAYING FOR CHILD CARE .................... Monthly Expenditures on Child Care ................ Monthly Expenditures, by Family and Child Characteristics ........................... Child Care Assistance .......................... Summary .................................. 5. PRESCHOOL ENROLLMENT .................. Care Settings for Preschool-Age Children ............. Characteristics Linked to Preschool Attendance.......... 1 5 5 6 8 8 9 10 13 13 14 18 23 27 29 29 30 32 36 39 40 43 xi Targeting Efforts to Groups Less Likely to Use Preschool ............................... Children in Parental Care Only .................. Children in Other Care Settings .................. Promoting Use of Preschool ...................... 6. COSTS OF UNIVERSAL PRESCHOOL............. Assumptions on Takeup of and Costs for Preschool ....... Families Eligible for Universal Preschool............. Enrollment of Children Not Currently in Preschool ..... Crowd-Out of Existing Preschool ................. Cost per Child in Universal Preschool .............. Alternative Estimates of the Costs of Universal Preschool ... Summary .................................. 7. CONCLUSIONS ............................ Appendix: Supplementary Tables ..................... Bibliography .................................. About the Authors ............................... Related PPIC Publications .......................... 50 50 52 53 55 55 56 57 57 59 60 64 65 67 73 75 77 xii Figures 3.1. Percentage of Children Ages 0–5 in Each Type of Care Arrangement ............................. 3.2. Percentage of Children Ages 0–5 in Each Type of Care Arrangement When All Parents Are Employed ....... 3.3. Percentage of California Children Ages 0–5 in Each Type of Care Arrangement, by Age ............... 4.1. Average Monthly Expenditure on Child Care by Parents Paying for Care ........................... 4.2. Payment Arrangements of Families Working and on Welfare ................................ 4.3. Use of Structured Care by California Families With and Without Child Care Assistance ................. 5.1. Child Care Settings for Children Ages 3–5 .......... 5.2a. Child Care Settings for Children Age 5 ............ 5.2b. Child Care Settings for Children Age 4 ............ 5.2c. Child Care Settings for Children Age 3 ............ 5.3. Share of California Children Ages 3 and 4 in Preschool, by Region ............................... 5.4. Share of California Children Ages 3 and 4 in Preschool, by Child Care Capacity in County of Residence ...... 5.5. Child Care Settings for Children Ages 3 and 4, by Household Composition/Employment ............ 5.6. Child Care Settings for Children Ages 3–5, by Number of Children Under Age 13 .................... 5.7. Child Care Settings for Children Ages 3–5 Not in Preschool ............................... 16 17 18 30 34 36 41 42 42 43 46 46 49 51 52 xiii Tables 2.1. Sample Characteristics of Children Ages 0–5 and Their Families ................................ 3.1. Percentage of Children Ages 0–5 in Child Care ....... 3.2. Percentage of California Children Ages 0–5 in Child Care, by Demographic Group and Type of Arrangement ............................. 3.3. Percentage of California Children Ages 0–5 in Child Care, by Work Schedule and Type of Arrangement .... 3.4. Hours and Number of Arrangements Used for California Children Ages 0–5 in Child Care, by Work Schedule ... 4.1. Monthly Child Care Expenditures for California Families Who Pay for Care, All Parents Employed ..... 4.2. Child Care Assistance for California Families Using Care for Children Ages 0–5, All Parents Employed ........ 5.1. Percentage of Children Ages 3 and 4 in Preschool, by Demographic Group ........................ 5.2. Regression Results on the Likelihood of California Children Ages 3 and 4 Being Enrolled in Preschool .... 5.3. Share of Preschool Arrangements That Are Part-, Moderate-, and Full-Time .................... 6.1. Assumptions on Takeup of Universal Preschool for California Children ......................... 6.2. Estimates of Average Annual Costs for Universal Preschool: High Estimate..................... 6.3. Estimates of Average Annual Costs for Universal Preschool: Low Estimate ..................... 6.4. Percentage of Costs Resulting from Participation of Currently Enrolled California Preschoolers, by Takeup Assumptions ............................. 6.5. Additional Slots Needed for California Children Enrolling in Preschool, by Takeup Assumptions ...... 7 14 20 24 25 31 33 45 48 53 59 60 61 63 63 xv A.1. Sample Characteristics for Family-Level Data (Costs) ... A.2. Percentage of Children Ages 0–5 in Rest of United States in Child Care, by Demographic Group and Type of Arrangement ....................... A.3. Counties Included in Each Region ............... A.4. Regression Results on Likelihood of Preschool Chldren Ages 3 and 4 Being Enrolled in California and Rest of United States ............................. A.5. Regression Results on Likelihood of Children Ages 3 and 4 Being Enrolled in Preschool in California ....... 67 68 70 71 72 xvi Acknowledgments This report would not have been possible without the extensive work of the Urban Institute’s Assessing the New Federalism project, their National Survey of America’s Families, and the assistance of their staff. Thanks as well to those who provided valuable comments including Deborah Reed, Shelley Waters Boots, Heather Rose, Susanna Loeb, and Laura Hill. Any opinions or errors are the authors’ alone. xvii 1. Introduction Child care policy in California reflects a struggle between two goals: supporting working parents and promoting child development. On the work support side, the 2000–2001 state budget included over $1.5 billion earmarked to help working families pay for child care, with the vast majority of funding going to families transitioning from welfare to work. The subsidies provided through this system reimburse parents for care provided by all but the most expensive providers. On the child development side, similar funding was provided through the Child Development Division of the California Department of Education (CDE) to support a variety of initiatives to promote access to highquality care, including support for state preschools, contracted slots in child development programs, services to increase the quality of care, and resource and referral services. Federal funding supports the Head Start program, which provides developmentally rich care for at-risk preschool children. As resources shrink in the state budget crisis, these two goals may come increasingly into conflict. The 2003–2004 budget made some cuts to child care subsidies; Governor Davis proposed a number of more dramatic changes, but they did not receive legislative support. These included deeper reductions in reimbursement rates, increases in family fees, and giving higher priority on waiting lists to families working fulltime. Other proposals called for moving control of child care subsidies from CDE to the Department of Social Services (DSS), which is more likely to focus on child care as a work support, or to the counties, where county flexibility could open the door to wide differences in county strategies to balance quality and affordability but could also put child care funding into competition with other county social service activities. Although such proposals seem to move the balance toward work support, other trends move it toward expansions in developmentally rich care, especially for children ages 3 and 4. Funded through a cigarette tax, 1 the California Children and Families Commission, also known as First 5 California, has made child development for children ages 0–5 one of its central goals. Two of its current initiatives alone provide over $500 million for school readiness programs and child development teacher retention. The Master Plan for Education, the David and Lucile Packard Foundation, and several county First 5 commissions, including Los Angeles and San Mateo, have embraced the call for universal preschool programs to increase child development and school readiness of children ages 3 and 4. In light of this rapidly changing policy environment, this report examines the current state of child care for young children in California, including the care arrangements currently used by families and the costs of this care. We then look more closely at issues from each side of the policy debate: first, the role of child care subsidies and other assistance in making care more affordable and, second, the issues involved in universal preschool. In particular, we address four key questions: • What share of California families use what types of child care for their young children, and how does this differ from the rest of the United States? • How much of their household income do families pay for child care and how much assistance do they receive in paying for it? • Are there major gaps between California and the rest of the country in preschool enrollment? If so, how much is explained by demographic differences? • If preschool access is expanded through universal preschool programs, how expensive will these programs be and how do different designs affect the cost of care through such programs? The analysis relies on 1997 and 1999 household survey data collected by the Urban Institute through the National Survey of America’s Families (NSAF). We consider a broad range of care settings: structured care in programs such as child care centers, nursery schools, preschools, and Head Start programs; family day care provided to a group of children in the provider’s home; care by relatives in the child’s home or the relative’s home; and nanny or babysitter arrangements 2 where an unrelated caregiver regularly provides child care in the child’s home. The report is organized as follows: Chapter 2 describes the National Survey of America’s Families, which provides the data for the report. Chapter 3 gives an overview of the level of child care use and the types of settings chosen by families with different demographic characteristics. Chapter 4 looks at the costs of care and the role of child care assistance, with some perspectives on the possible effect of proposals to change subsidy policies. Chapter 5 assesses preschool enrollments and examines the issues involved in promoting greater enrollment. Chapter 6 then addresses the effects of different assumptions on the estimated cost of universal preschool. We offer conclusions in Chapter 7. 3 2. Data The National Survey of America’s Families The data for this study are drawn from the National Survey of America’s Families, a multiyear survey conducted by the Urban Institute as part of its Assessing the New Federalism project.1 The survey was designed to learn more about the economic, health, and social circumstances of children and adults under age 65; it includes a representative sample of the U.S. non-institutionalized, civilian population with oversampling of low-income families, as well as additional observations in 13 target states, including California. Our study uses data from the first two survey rounds of the NSAF: 1997 and 1999. In 1997, 2,543 households in California were interviewed out of 44,000 households nationwide. The 1999 survey includes 2,206 California households out of over 42,000 interviewed nationwide.2 Interviews were conducted by telephone, and to capture households without telephones, interviews were conducted using cellular phones loaned to these households. In households with children under age 18, up to two children could be chosen for in-depth study: one child under age 6 and one child age 6–17. Interviews about these children were conducted with the adult in the household who was the most knowledgeable about each sampled child. In California, 95 percent of these adults were the biological parents. Eighty-one percent were female, of which most (91 percent) ____________ 1Information about Assessing the New Federalism can be found on the Urban Institute website at www.urban.org. 2Additional information on the design of the NSAF can be found on the Urban Institute website at http://newfederalism.urban.org/nsaf/methodology1997.html for 1997 and http://newfederalism.urban.org/nsaf/methodology1999.html for 1999. 5 were the biological mothers. Nineteen percent were male, of which 93 percent were the biological fathers.3 Sample of Families with Young Children These analyses of child care focus on families with a child age 0–5 that were interviewed during the school year in 1997 and 1999.4 Further, we compare eligible families living in California to similar families living in the rest of the United States. To retain an adequate number of individuals in our sample for California for this comparison, we combined the interviews conducted in 1997 and 1999. After combining the observations, our sample for California includes 1,287 children in 1,282 families. For the rest of the United States, our sample includes 20,350 children ages 0–5 in 20,328 families. Sample weights were adjusted to account for the different sample and population sizes in the two years.5 Families with young children in California differ substantially from families with young children in the rest of the nation. Table 2.1 presents an overview of the characteristics of the children and their parents. In particular, almost half of California children ages 0–5 are Hispanic compared to only 13 percent nationwide. California children are less likely to be African American and more likely to be Asian. However, the sample sizes for each of these two groups are too small to support subgroup analysis of the California sample. Therefore, we will focus only on differences between non-Hispanic whites and Hispanics in any breakdowns by race/ethnicity. Racial and ethnic makeup is not the only significant difference between California families and those living elsewhere in the nation. ____________ 3For simplicity, we will use the term parent(s) henceforth. 4Child care arrangements vary substantially between the summer months and other seasons (Cappiziano, Adelman, and Stagner, 2002). Thus, we focus our analyses on care occurring in non-summer months. The sample size is too small to allow examination of summer child care use in California. 5These weights adjust the sample weights developed by the Urban Institute to balance across the two years and scale up to an average of the population between 1997 and 1999. Discussion on the creation of the underlying NSAF weights can be found in the methodology reports at http://newfederalism.urban.org/nsaf/methodology.html. 6 Table 2.1 Sample Characteristics of Children Ages 0–5 and Their Families Characteristics of Children and Parents Child’s age 0–2 3–5 Child’s race/ethnicity Non-Hispanic white Hispanic Other Income as a % of federal poverty limit Less than 100 100–199 Over 200 Most knowledgeable adult’s welfare history Currently on welfare Formerly on welfare Never on welfare Most knowledgeable adult’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Most knowledgeable adult’s marital status Single parent Married or cohabiting parent Adults 18 or older in household in addition to parents Other adults in household No other adults in household California Weighted No. % 576 49 711 51 443 38 652 45 192 17 361 27 343 22 583 51 156 12 154 11 970 77 270 21 358 26 296 22 350 31 309 21 978 79 297 22 990 78 Rest of United States Weighted No. % 9,156 11,194 49 51 13,498 2,985 3,867 67 13 20 4,320 5,383 10,647 21 24 56 1,230 3,048 15,976 6 13 81 2,231 6,402 4,734 6,870 11 31 23 35 5,169 15,181 22 78 3,229 17,121 14 86 Children under age 6 living in families in California tend to be poorer than similar families in the rest of the United States. The most knowledgeable adults (typically mothers) interviewed about these children are also less educated, with twice as many not completing high school (21 percent compared to 11 percent). In the 1997–1999 period, welfare participation is also higher in California. Almost twice as many young children in California live with adults who are currently on 7 welfare (12 percent compared to 6 percent). Family structure for these children differs somewhat in California. Although single parenthood is about equally common in California as in the rest of the county, California children are more likely to live in extended families. Analysis Variables Child Care Arrangements The telephone interview included questions about the regular use of specific types of child care for the selected child during the past month. Regular use was defined as at least once a week during the past month. In addition, the adult respondents were asked a number of questions about the care used, including the number of hours the arrangements were used, whether the caretaker was a relative, and whether the arrangement was used while the respondent, usually the parent, worked, went to school, or participated in job training. Adult respondents were also asked whether the selected child attended kindergarten.6 From this information, we determine whether the selected child regularly participated in non-parental care in any of the following settings: center care, Head Start, nursery school, kindergarten, before and after school care, family day care, nanny or babysitter care, in-home care by a relative, and out-of-home care by a relative. When parents used none of these non-parental care arrangements, the child’s care setting is designated as parental care. In some cases, parents may take their children to work or work back-to-back schedules so that they can care for their children themselves. In other cases, the respondents may use irregular arrangements that vary from week to week or they may leave their children unattended but are reluctant to admit this in the interview. These cases are included in the parental care category, although the extent of their prevalence is unknown. We first constructed a four-category employment variable by household composition: single parent, employed; two parents, both employed; two parents, one employed; single or two parents, no one ____________ 6Because our data did not include birth dates, we cannot say how many of the children age 5 in our sample were eligible for kindergarten. 8 employed.7 We then divided households into two employment categories: fully employed and not fully employed. A household is fully employed if all parents are in the workforce regardless of the number of hours worked. A household is considered not fully employed if at least one parent is not in the workforce. Child Care Expenses and Assistance The NSAF asked parents how much they paid out-of-pocket for child care arrangements8 in the last month for all their children under age 13 while they worked, were in school, or looked for work.9 Thus, our analyses of child care expenses and the receipt of child care assistance refer to the costs incurred by the families with working parents for child care for all their children under age 13.10 We focus our analyses only on families who had at least one child age 0–5 and who used at least one child care arrangement while a parent was employed. Note that some families who had at least one non-employed parent also reported using non-parental child care arrangements. However, the survey did not collect information about the costs incurred for this care. We coded working parent families as receiving assistance in paying for child care based on their responses to a number of questions posed by the survey. All families that used child care for employment purposes were asked if anyone helped to pay for all or part of the cost of care, regardless of whether they reported any child care expenses. If they answered yes, they were asked to indicate who or what agency helped them pay for care. We focused our analyses on reports of assistance from ____________ 7Two-parent households include married parents as well as a parent and a cohabiting partner. 8It is not possible to examine the costs of each child care arrangement separately or to examine the aggregate costs of each child in the family’s care arrangement. 9If the two most knowledgeable adults lived in the same family (for example, each spouse or unmarried partner was most knowledgeable about a different focus child), the cost data each reported were combined and the family was treated as one case (Giannarelli, Adelman, and Schmidt, 2003). 10Appendix Table A.1 provides a breakdown of family characteristics in our cost sample. 9 three general sources: help from the government, help from a relative, and other help.11 In addition, if families did not report receiving help but they indicated that the amount they paid for care was determined by how much they earned, we coded them as receiving help from the government.12 Families that used child care while they worked and did not report receiving help but did not report any child care expenses were assumed to be receiving “free” child care. They were assigned to one of three categories of assistance depending on their child care arrangement. If they used structured care, we assumed that they were receiving free government help. If a relative provided child care, we assumed that they were receiving free relative care. And if they were in some other type of arrangement, such as using a nanny/babysitter or family day care, we assumed that they were receiving free other care. Cases that were in more than one type of arrangement were counted in more than one assistance category. For example, a family that reported no child care costs but that used a child care center part-time and care by a relative part-time would be included in both the free government help and free relative care categories. Economic and Demographic Variables Throughout this analysis, we compare the child care arrangement and cost patterns for different groups of children and families based on their economic and demographic characteristics. Although most of the categories are straightforward, several are explained below. ____________ 11Our analyses of child care assistance follow the Urban Institute’s approach to differentiating the sources of assistance (see Giannarelli, Adelman, and Schmidt, 2003). Some families who were designated as receiving “other” assistance may in fact be receiving a government subsidy. We cannot distinguish with certainty the sources of assistance for families who are in a private child care program and receiving assistance. The subsidies that some of these families are receiving may be funded through a government program. For this reason, government assistance may be underestimated (Giannarelli, Adelman, and Scmidt, 2003). 12Some of these families could in fact be paying on a sliding scale with a private organization that is not reimbursed by the government for this assistance; therefore, “other” assistance may be underestimated. 10 In addition to poverty status, we categorized a family’s annual income relative to the State Median Income (SMI). For all families, we divided their income by California’s median income for their respective family size and interview year. We chose to use this categorization because income relative to the SMI determines which families are eligible for subsidized child care in California, our focus state. We constructed a measure of welfare history using survey data on current and past participation in welfare (Temporary Assistance for Needy Families, TANF, or Aid to Families with Dependent Children, AFDC). Using these data and the SMI ratio, we constructed a welfare history variable with four categories: (1) currently on welfare, (2) formerly on welfare (this included families that had ever received welfare), (3) never received welfare but income-eligible for subsidies, and (4) higher income and never received welfare.13 ____________ 13The third and fourth groups are distinguished by whether their income was above or below 75 percent of California SMI, the eligibility cutoff at the time of the surveys. 11 3. Overview of Child Care Use in California The complexity of child care policy arises from the complexity of child care itself. Parents use non-parental care to allow them to work and to provide a socially and developmentally enriching environment for children, especially as children approach school age. In this chapter, we provide an overview of the use of child care in California and in the rest of the nation, including the arrangements chosen by parents—ranging from formal Head Start programs to informal care provided by relatives in the children’s homes—and the characteristics of children and families linked to different choices of care. Share of Children Under Age 6 in Child Care Sixty-five percent of California children under age 6 were regularly placed in at least one child care arrangement in a typical month between 1997 and 1999. As shown in Table 3.1, this rate was significantly lower than the rate of child care use for children in the rest of the nation, where 73 percent were in care. This difference is partly driven by the lower labor force participation of parents of young children in California. Fifty-four percent of U.S. children live in households where all parents are employed, compared to only 46 percent in California, primarily because of lower labor force participation of mothers in two-parent households. (Labor force participation among single parents is also lower in California.) However, child care use is lower in California regardless of employment status or household structure. We distinguish between single-parent and two-parent families and between families in which all parents are employed or not (both employed in the case of two-parent families or one employed in the case of one-parent families). Single parenthood is about as likely in California as in the rest of the 13 Table 3.1 Percentage of Children Ages 0–5 in Child Care All children Parent(s) employed Single parent, employed Two parents, both employed Parent(s) not employed Single parent, not employed Two parents, one or both not employed California 100 46* 10* 36* 54* 11* 43* *Indicates significance at the 95 percent level. Rest of United States 100 54 14 41 46 8 38 % of Group in Non-Parental Care Rest of United California States 65* 73 83* 88 89 92 81 87 49* 54 62 64 45* 52 United States, but single parents are less likely to work in California. Similarly, two-parent families in California are less likely to have both adults in the labor force. Child care use is highest among working single parents and lowest among two-parent families with one or both parents out of the labor force. For each of the four family types, use of child care is lower in California than in the rest of the United States, although within these groups the California-U.S. difference in care use is significant only for two-parent families with one or both adults not employed. Although child care is frequently viewed as a support for working parents, the majority of young children in the United States spend time in non-parental care even when there is at least one parent not working, in school, or in training. The share is lower in California, but even in California nearly half of young children with non-working parents are in at least one care arrangement. In large part, this reflects the importance of child care arrangements in supporting child development, an issue we return to in Chapter 5. Choice of Arrangements Child care is offered in a number of different settings. The grouping of these arrangements depends on the question to be considered. In this 14 section, we are interested in the options selected by families—the result of decisions balancing their preferred modes of care, the care available in their community, the costs of various settings, and the convenience and quality of the alternatives. In considering the options parents selected, we group settings into five major categories: Structured care represents the most formal arrangement, including care at both centers and schools. This care is most likely to be licensed and, typically, the most consciously oriented toward child development. Under this category, we include child care centers, nursery schools, preschools, and Head Start programs. In addition, some children age 5 are enrolled in kindergarten.1 Although kindergarten may be thought of as school rather than child care, it is optional in California. The state also allows younger children into kindergarten than do many other states. Because California families may have greater opportunities to substitute kindergarten for other developmentally rich child care arrangements, we include kindergarten in the structured care category.2 Family day care is care provided to a group of children in the provider’s home. These providers may or may not be licensed, depending on state regulations. In California, providers who care for children from more than one family (other than their own) must be licensed. However, the NSAF data do not distinguish between licensed and unlicensed family day care. Relative care is divided between care provided by relatives in the child’s home and care provided by relatives in the relative’s home. Extended family members, such as grandparents, have always been an important and trusted source of care for young children. Such care when provided in the child’s home is often given by extended family living in the same household, whereas such care outside the child’s home could ____________ 1Surveys were completed between February and November, so some children age 4 could have been eligible for kindergarten at the time of the fall surveys. However, we did not observe any of them in kindergarten. 2Earlier Urban Institute analyses of care options excluded kindergartners. However, given that center care is relatively less common for children under age 6 in California, and kindergarten attendance is somewhat more common, we believe that the inclusion of kindergartners in structured care more accurately reflects the greater use of such care in California than in the rest of the United States. 15 range from a very informal arrangement with a grandmother to a businesslike arrangement that differs little from family day care. Thus, parents may or may not have to pay for care given by a relative. Nanny or babysitter care represents the final category in the analysis below. These are arrangements where an unrelated caregiver provides child care in the child’s home.3 We do not distinguish between nannies and babysitters, although only regular ongoing arrangements are included. Figure 3.1 shows the share of young children cared for regularly in each of these five categories.4 70 60 *58 *52 50 California Rest of the United States Percentage 40 30 22 21 25 25 †27 †23 20 14 12 10 0 Structured care Family day care Care by Care by relative Nanny/ relative in outside babysitter child’s home child’s home NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. † Indicates significance at the 90 percent level. Figure 3.1—Percentage of Children Ages 0–5 in Each Type of Care Arrangement ____________ 3Nanny care offered in another person’s home is generally not distinguishable from family day care. 4This categorization of care settings differs from that reported in Urban Institute publications in two ways. The Urban Institute typically reports only the primary care setting (the setting in which a child spends the most hours) and concentrates on care used while parents are employed. 16 Structured care is the most common setting for children under age 6. In California, 52 percent of young children in some care arrangement spend time in a child care center, nursery school, preschool, Head Start, or kindergarten during a normal week. Care by relatives is the next most common setting. In-home and out-of-home care by relatives combined accounts for nearly as many children as structured care. California families do not differ substantially from other American families in the type of child care used, except that young children in California are less likely to be in structured settings. Care by relatives outside the home is also somewhat less common for California children, and care by nannies or babysitters is slightly more common. The share of children cared for in family day care homes, in California and elsewhere, rises if we restrict our view to families where all parents are employed (Figure 3.2). This shift to family day care is somewhat more pronounced for California children. 70 60 *57 50 *50 California Rest of the United States Percentage 40 30 27 24 *24 *26 25 28 20 15 12 10 0 Structured care Family day care Care by Care by relative Nanny/ relative in outside babysitter child’s home child’s home NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 3.2—Percentage of Children Ages 0–5 in Each Type of Care Arrangement When All Parents Are Employed 17 Choice of Arrangements, by Child and Family Characteristics Parents’ choice of arrangements for their young children is a complex decision, but clearly characteristics both of the children and their families are important. The age of the child is perhaps the most important characteristic for understanding the choice of arrangements. Parents of very young children, those under age 3, are less likely to place their children in care. Even though 65 percent of all young California children are in some care arrangement, only 52 percent of children ages 0–2 are in care, compared to 76 percent of children ages 3–5. One in five infants and toddlers with working parents are still cared for by parents only, meaning that work schedules must be arranged to keep a parent at home with the child. In contrast, structured care in California is primarily used for children ages 3–5, as we see in Figure 3.3. Without conditioning on a child being in care, infants and toddlers are more likely than older children to be cared Percentage 90 80 *76 70 60 50 40 30 20 *19 10 0 Structured care Children Ages 0–2 Children Ages 3–5 *37 *18 *31 *18 *25 *18 17 12 Care by Care by relative relative in outside childs home child’s home Family day care Nanny/ babysitter NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 3.3—Percentage of California Children Ages 0–5 in Each Type of Care Arrangement, by Age 18 for by relatives. Among children in care, all arrangements other than structured care are more common for the youngest children. The age of the child, however, is not the only demographic characteristic that is important in understanding choice of care arrangements. Table 3.2 shows the choice of arrangements for young children in California disaggregated by six family characteristics: family structure, child’s race/ethnicity, parental education, the presence of other adults in the household, income, and welfare receipt.5 Income and education are both important factors in characterizing which families use child care. The first column of Table 3.2 shows the share of children in each group in at least one non-parental arrangement, regardless of parents’ employment status. Disadvantaged families are the least likely to use any child care. Families with income below poverty are least likely to use non-parental care. As income rises, the share of families using care also rises. The share using child care also rises with education. Fifty-five percent of parents with less than a high school diploma use child care for their young children, compared to 73 percent of parents with at least an associate degree. Families with less income and less education also have lower labor force participation, but the differences in child care use hold even if we consider only working families. One of the most striking differences in child care use is for Hispanic children. Forty-one percent of these children are cared for by parents only, compared to 29 percent of non-Hispanic white children. However, this difference can be largely explained by other factors. Hispanic children are more likely to live in two-parent families, with lower average income, lower average education, and larger family sizes. Controlling for these characteristics, there is no statistically significant gap between Hispanic and non-Hispanic families in the likelihood of using child care for their young children. Hispanic families who do use child care for their young children are less likely to select structured care settings and more likely to select relatives as caretakers. Not surprisingly, the presence of other adults in the household has a large influence on the care choices of parents. Fortyseven percent of young children with another adult in the home are cared ____________ 5Appendix Table A.2 shows a similar breakdown for the rest of the United States. 19 Table 3.2 Percentage of California Children Ages 0–5 in Child Care, by Demographic Group and Type of Arrangement % in NonParental Care % in Structured Care by Relative Care in Child’s Home % in Arrangementa Care by Relative Outside Child’s Family Home Day Care Nanny/ Babysitter 20 Family structure Single parent Two parents Child’s race/ethnicityb Non-Hispanic white Hispanic Parental educationc Less than high school High school diploma/GED Some vocational education/college Associate degree or more Other adults in household No Yes Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI 75* 62 71 59* 55* 58 69* 73* 65 64 52* 65* 70 73 60 32 51 23 56 18 46* 34* 47 33 48 28 54 26 59* 21 54 18 47* 47* 55 25 47* 32* 51 28 58 19 20 22 8* 24 21 15 20 25 20 27* 17 8* 24 13 7 28 21 5 26 26 14* 18 23 21* 25 23 15 16 16* 3* 21 18 11* 28 18* 10* 23 22 2* 21 26 21 21 Table 3.2 (continued) Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % in NonParental Care % in Structured Care by Relative Care in Child’s Home % in Arrangementa Care by Relative Outside Child’s Family Home Day Care 72* 56 21 55 49 33 73* 55 21 59* 49 32 21 25 25 16 27 22 19 17 NOTE: Italics indicate reference group for significance tests. aChildren may be in more than one arrangement. bSample sizes in California are too small to report breakouts for black and Asian children. cBased on years of education of the “most knowledgeable adult” for the focal child. *Indicates significance at the 95 percent level. Nanny/ Babysitter 17 11 8 8 for by a relative in the home, compared to 18 percent of children without other adults in the home. The households of Hispanic children are much more likely to include other adults (27 percent, compared to 18 percent of households of non-Hispanic children), explaining the high use of relatives in the home by this group.6 Hispanic children, however, are also more likely to be cared for by a relative outside the home. The choice of child care arrangement significantly differs by parents’ level of education: Parents with more education are more likely to use structured care and nannies or babysitters. Similarly, an increase in income does not simply predict choice of care arrangement. The income categories in Table 3.2 are based on two conceptual threshold levels of income: the poverty threshold and the SMI. At present, families with income below 75 percent of the SMI are eligible for subsidized child care in California. The SMI measure in effect from September 1998, created by the California Department of Finance, places this median at $50,148 for a family of four. This means that the 75 percent cutoff fell at approximately $37,611—about 200 percent of the current federal poverty threshold of $18,400.7 As noted above, lower-income families are less likely to have their children in care, but there is no arrangement whose use rises consistently with income. For example, poor families are about as likely to use structured care as families with income above the state median income. Programs targeted to the poor may explain much of this similarity. Families just above the poverty threshold are less likely to use structured care. The opposite pattern holds for care by relatives: families with income low enough to qualify for subsidies but above the poverty threshold are more likely to rely on relatives than either higher- or lowerincome families. Families with income above the state median are most likely to have nannies or babysitters, but families in the next highest income category are the least likely. ____________ 6Other research on ethnic differences in child care selection found that Hispanic parents were significantly less likely than white parents to choose center care; however, after controlling for the presence of other adults in the home, the lower propensity of Hispanics to use center care diminished (Liang, Fuller, and Singer, 2000). 7The SMI cutoff has not been updated for inflation. 22 Welfare receipt has a curious connection to child care use. The last section of Table 3.2 disaggregates families into four groups based on welfare history and eligibility for subsidies: never received welfare and income too high to qualify for subsidies, never received welfare and income low enough to qualify for subsidies, did not receive welfare at the time of the survey but previously received it, and receiving welfare at the time of the survey. Former welfare recipients have both child care usage and arrangement types that are very similar to higher-income families. They have high child care usage, are most likely to use structured care, and are somewhat more likely to use family day care. Former welfare families, who typically transitioned from welfare to work, are eligible for child care assistance. Former welfare recipients’ higher use of licensed care and particularly center care is consistent with findings from administrative data on child care subsidies and surveys of subsidized child care settings (Marrufo, O’Brien-Strain, and Oliver, 2003).8 On the other hand, families currently on welfare have patterns very similar to those of low-income families who have never been on welfare. These families have lower use of any child care, and they are more likely than other families to rely on care by relatives, especially in the child’s home. Current welfare recipients also qualify for child care assistance, but only 26 percent of current recipient families have all parents working, compared to 56 percent of former welfare recipients. In the next chapter, we explore further the role of child care subsidies. Work Schedules and Care Arrangements About half of young children with a parent at home spend time regularly in a non-parental setting, as shown in Table 3.3. The majority of these children are placed in structured care settings, although other arrangements are also used, especially care provided by relatives. In families where all parents work (a working single parent or two parents who both work), work schedules influence both the probability of using non-parental care and, to a lesser extent, the type of arrangements chosen. Most notably, parents who work during evenings ____________ 8The higher use of structured care among former welfare recipients may result because former recipients switch their use of care as their work and household stabilize. 23 Table 3.3 Percentage of California Children Ages 0–5 in Child Care, by Work Schedule and Type of Arrangement 24 At least one parent not employed Parents employed At least one part-time, only regular hours Full-time, only regular hours At least one part-time, some evening or weekend hours Full-time, some evening or weekend hours % in NonParental Care 49* 83 87 61* 81 % in Structured Care 58 45 54 32 47 NOTE: Italics indicate reference group for significance tests. aChildren may be in more than one arrangement. *Indicates significant at the 95 percent level. % in Arrangementa Care by Relative Care by Relative Outside Child’s in Child’s Home Home 24 20 Family Day Care 13* Nanny/ Babysitter 11* 29 24 26 26 21 22 33 12* 37 24 16 28 33 32 17 11 or weekends are less likely to rely on non-parental care and most likely to use care by relatives. Use of structured care is not notably lower for these families, except for those working non-standard hours part-time. Babysitters and nannies are more commonly used by families where both parents are employed but not full-time. Reflecting the different uses of child care, the number of hours per week that children spend in care differs between families where all parents work and families with at least one parent at home. Not surprisingly, children with non-working parents spend much less time in care than do those with all parents in the workforce (Table 3.4). On average, a child with working parents spends 35 hours per week in care, whereas children with a parent at home spend 18 hours per week in care. Children whose parents work full-time during regular hours spend more time in non-parental care than do children whose parents work only part-time or non-standard schedules. This, coupled with the lower Table 3.4 Hours and Number of Arrangements Used for California Children Ages 0–5 in Child Care, by Work Schedule All children Children with At least one parent not employed Parents employed At least one part-time, only regular hours Full-time, only regular hours At least one part-time, some evening or weekend hours Full-time, some evening or weekend hours Parents employed Single parent Two parents Average Hours in All Arrangements 28 18* No. of Arrangementsa 3 or 1 2 More 60 32 8 66* 30 4 27 53 38 8 38* 54 34 12 23 60 33 7 31 60 30 11 41* 40* 43 17 32 61 31 9 NOTE: Italics indicate reference group for significance tests. aTotals may not sum to 100 because of rounding. *Indicates significance at the 95 percent level. 25 number of hours in care for children from two-parent families, demonstrates the role of reduced or offset schedules as a strategy for minimizing the time children spend in care. Children of working single parents spend an average of 41 hours per week in care, while those with two working parents spend only 32 hours per week in care on average. This difference results partly because of the shorter work hours common for second earners and partly because parents can vary their schedules to minimize the hours of child care needed. Across all these groups, the hours in care are virtually identical for California families and other American families. Overall, 40 percent of California children in care spend time in two or more different arrangements in an average week, as shown in Table 3.4. (Note that the figures in Table 3.2 are per child; thus, the complexity of care for families with multiple children is understated.) In some cases, the use of multiple arrangements appears to be a choice, probably related to a balance between developmentally rich care and affordable care, suggested by the fairly high use of multiple arrangements (34 percent) even among families with a parent not in the workforce. However, the need for multiple arrangements is also closely tied to work schedules: Working single parents, who lack a backup person to care for their children while they work, not only have their children in care for the longest hours but are also most likely to depend on multiple arrangements. This is particularly true of single parents in California, where 60 percent of single parents rely on multiple arrangements, compared to only 49 percent of single parents elsewhere in the United States. Although this information is not included in the table, the presence in the household of other adults, beyond the parents, counteracts this effect, where children with additional adults at home are least likely to be in multiple care arrangements when their parents work. Disadvantaged families are more likely to require multiple arrangements to facilitate work: Low-income working families and families combining welfare and work are the most likely to rely on multiple arrangements for very young children (under age 3). 26 Summary Fewer young children are placed in child care in California than elsewhere in the United States. Still, most children of working parents (83 percent) regularly spend time in non-parental care, averaging 35 hours per week. Children of working single parents spend the most time in care and are also the most likely to be in multiple care arrangements. However, child care is not used only to support work. Forty-five percent of children with a parent not in the labor force are also regularly placed in non-parental care, albeit for fewer hours per week than the children of working parents. Certain families are less likely to use care. In particular, children with less-educated parents and poorer children are less likely to be placed in care. Hispanic children are also less likely to be placed in care, although this difference is explained by the lower average education and lower average income for this group. Children under age 3 are less likely to be placed in care than are preschool-age children, and when they are in care, they are more likely to be cared for in an informal arrangement such as by relatives. Children with adults in the home other than their parents are also most likely to be cared for by relatives. Higher-income families are more likely than lower-income families to place children in structured care settings, although families formerly on welfare have patterns of child care use very similar to those of higher-income families. Finally, although multiple care arrangements are common even for children with parents out of the labor force, the ability to coordinate work schedules has a strong influence on the number of arrangements used for young children, as well as on their hours in care. Working single parents use both the most hours of care and the largest number of settings for their young children; families with some non-standard hours can limit both the number of hours in care and the number of arrangements. However, the coordination of part-time work and nonstandard hours seems to draw children away from structured care settings, with greater reliance on relatives when care is used. 27 4. Paying for Child Care Public policies concerned with child care typically focus on helping families pay for care. In California, child care assistance comes in a variety of forms, ranging from tax credits for all families—now available as a refundable credit—to targeted subsidies paid essentially as vouchers for current and former welfare recipients. In this chapter, we examine families’ expenditures for child care and the degree to which they receive assistance, from government or other sources, in paying for care. The NSAF captures information on the cost of child care in a different structure from the information it collects on child care arrangements. Whereas arrangement information is based on observations for a focal child, the cost of child care is captured at the family level for all children under age 13. Child care costs for all children in the family more accurately reflect the role of child care costs in a family budget and the effectiveness of assistance in supporting lowincome families. A major drawback is that we cannot relate costs directly to the care arrangements used. In addition, we cannot identify the specific arrangements for which families are receiving assistance. Monthly Expenditures on Child Care Among families who pay for child care, the average California family with young children spent $360 per month out-of-pocket on child care for all children in the family (Figure 4.1). As expected, given the differences in hours of care, families with at least one parent out of the labor force pay less for care, averaging $296.1 Families with both parents employed (or an employed single parent) pay $373 per month for care. Whether or not all parents are employed, California families on average pay more than families elsewhere for child care. However, because ____________ 1Families with no out-of-pocket costs are not included in the average cost calculations. 29 Expenditure ($) 400 350 300 250 200 150 100 50 0 All families who pay for care California Rest of United States Families with all parents employed Families with parents at home Figure 4.1—Average Monthly Expenditure on Child Care by Parents Paying for Care California families had higher earnings on average, child care costs were actually a slightly smaller share of earnings: 10 percent in California, compared to 11 percent elsewhere. Child care policies that help pay for care, as opposed to providing free care through Head Start or state preschools, typically require that parents be working to qualify for aid. This support is generally designed to make work more affordable, whether provided through the welfare system, the non-welfare subsidy system, or the tax system. For this reason, the rest of this chapter focuses on families with all parents employed. Monthly Expenditures, by Family and Child Characteristics Since monthly expenditures are reported by family rather than by child, we would expect child care expenditures to vary by the number and ages of children served. Indeed, families that include a child under age 3 typically had to spend about $36 more per month on child care than families in which the youngest child was age 3–5 (Table 4.1). However, families with two or more children under age 13 (one of 30 Table 4.1 Monthly Child Care Expenditures for California Families Who Pay for Care, All Parents Employed All working parent families Child’s race/ethnicity Non-Hispanic white Hispanic Other adults in household No Yes Age of youngest child 0–2 3–5 Number of children under age 13 1 2 or more Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % Paying for Care 72 Expenditures Mean Median As % of ($) ($) Earnings 373 320 10 75 430 400 11 69 273 240 11 77 383 325 10 55 324 311 13 71 392 332 11 74 356 320 10 73 365 325 10 71 379 311 11 57 242 194 24 67 283 300 13 75 350 268 9 73 417 400 7 76 426 400 7 66 262 249 13 69 309 260 13 59 414 337 29 whom is under age 6) do not pay significantly more than families with only one child requiring child care. This may reflect different strategies for families with several children or the presence of older siblings to help care for younger children. We do see that families with other adults in the household pay less for care, when they do pay for care, than families with no other adults. Hispanic families also pay much less for child care than do non-Hispanic white families, even after controlling for differences in income and the presence of other adults in the household. However, child care as a share of earnings is very similar for the two racial/ethnic groups. 31 Family income is the most important determinant of monthly expenditures on child care. Families with income above the state median paid an average of $417 per month on child care, compared to $242 paid by poor families. However, despite paying much less in absolute terms, lower-income families pay much more as a share of earnings. Poor families who do pay for care pay an average of 24 percent of earnings, compared to about 7 percent for families with income above the state median. Working families who are also receiving welfare pay the highest child care costs as a share of earnings. As we discuss below, most current welfare families receive assistance with child care, but among those who pay for care, the reported monthly expenditure of $414 on average is similar to that paid by high-income families. However, this appears to be driven by a few families paying very high amounts (and perhaps getting reimbursed), since the median monthly expenditure, although still higher than that for other low-income families, is much closer to that for the typical family.2 Child Care Assistance Many parents whose young children are in child care while the parents are working receive assistance in paying for this care (Table 4.2). Twenty-eight percent pay no out-of-pocket costs, because the care is provided free by relatives, subsidized by the government (through either subsidy vouchers, state supported preschools and kindergarten, the Head Start program, or other government or social service assistance), or paid by employers or absent parents.3 This is true even for higher-income families: 25 percent of families with income over 75 percent of the state median income paid no out-of-pocket expenses for child care. Families with other adults in the household were the least likely to pay for care. In California, only 55 percent of working families with other adults in the household paid for any care arrangements. One in ____________ 2It is possible that some parents included reimbursed cost when they reported their out-of-pocket costs. 3Families may also receive a dependent care tax credit, but these credits are not included in NSAF questions about child care assistance. 32 Table 4.2 Child Care Assistance for California Families Using Care for Children Ages 0–5, All Parents Employed All employed parent families Child’s race/ethnicity Non-Hispanic white Hispanic Other adults in household No Yes Age of youngest child 0–2 3–5 Number of children under age 13 1 2 or more Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % Paying No Child Care Costs 28 % Receiving Help Paying for Care Government Free Any or Social Care by Assistance Services Relative 37 19 14 25 31 31 44 12 12 26 14 23 32 45 56 17 11 26 24 29 33 26 41 15 13 23 14 27 37 29 36 43 74 33 46 25 43 27 28 17 16 21 12 46 14 28 10 18 13 11 12 24 25 34 51 31 54 41 68 8 14 27 16 33 14 66 6 four of these families received free care from relatives, nearly twice the rate of help from relatives for all working families in California. Free care by relatives was not the only source of assistance for families with other adults in the household: Nearly as many received subsidies or other assistance from the government or social service agencies.4 ____________ 4Care by relatives can be reimbursed through the welfare system or the non-welfare subsidy system. 33 One-third of families with income below 75 percent of the state median and 43 percent of poor families reported that they paid no outof-pocket expenses for child care. Almost three-quarters of poor families reported that they received some assistance in paying for care, including 46 percent receiving assistance from government or social services. This rate of government assistance is higher than the 36 percent rate for such families in the rest of the United States. More than one-fourth of California families with income above poverty but below 75 percent of the state median income also receive government help. This rate is much higher than elsewhere in the United States, because families in this range elsewhere are commonly above their local cap for federally funded assistance. The government assistance category, however, includes families that do not specifically report subsidies but are using free care through preschools or Head Start or other organized settings, so this may also partially reflect the continuing role of federal- and state-funded care in California. Welfare families are another group particularly likely to receive assistance with child care. In fact, in California, as in most states, working families on or recently off welfare are eligible for child care assistance through the TANF program. As Figure 4.2 shows, families 70 California 60 Rest of United States 50 Percentage 40 30 20 10 0 Pay for care Receive government Receive free care social services assistance from relative Figure 4.2—Payment Arrangements of Families Working and on Welfare 34 combining work and welfare in California are both more likely to receive government help paying for child care and more likely to pay some outof-pocket expenses. In fact, nationally only 36 percent of working families currently on welfare report paying for their child care, compared to 59 percent of those in California. But 66 percent of families report receiving government help in California, compared to 53 percent elsewhere. In California, in fact, it appears that there may be some tradeoff between government help and help from relatives, since only 6 percent of families currently on welfare report receiving free child care from relatives, less than one-third the share elsewhere. Surprisingly, child care assistance appears to often augment rather than offset spending on child care. Families with child care assistance are less likely to pay any out-of-pocket costs, but many low-income families receive help but still report paying for care. Families that pay for care while receiving assistance pay about the same amount per month as families with similar income but no assistance. For example, California families with working parents and income below 75 percent SMI paid on average $277 per month whether or not they received help with care. The effect is even more striking for welfare families: Those families currently on welfare and receiving assistance—when they pay for care— spend more than paying families without assistance: $460 per month compared to $395 per month. Because even among low-income families, those receiving help are likely to have lower earnings than those without help, families with assistance who still face costs are actually paying a larger share of their earnings for child care than families without assistance. In fact, low-income families may be using child care assistance not simply to lower the costs of care but also to allow them to select more developmentally oriented settings for their children. Among families with young children, low-income working parent families with assistance are far more likely to have their children in structured care settings than are such families without assistance.5 This is particularly true of families ____________ 5Research by Hirshberg, Huang, and Fuller (2002) of parents moving from cash aid to work activities in three California counties suggests that although subsidies may increase the likelihood of choosing center care, ethnicity, marital status, and home 35 currently receiving welfare: 68 percent of welfare families with assistance use structured care, compared to only 16 percent without assistance (Figure 4.3). On the other hand, the link between structured care and assistance may go the other way. That is, even though many forms of child care assistance, especially for welfare families, can be applied to all child care settings, it may be that assistance is much more closely tied to organized settings (through Head Start or state-subsidized preschools) or that families choosing organized care are much more likely to seek assistance than are families using other settings. 80 Families without assistance 70 Families with assistance 60 Percentage 50 40 30 20 10 0 Families with all parents employed Low-income families Families currently on welfare Figure 4.3—Use of Structured Care by California Families With and Without Child Care Assistance Summary Among families who use child care while both parents are working (or a single parent is working), 72 percent pay some out-of-pocket costs, averaging $373 per month. This cost is higher than the average in the rest of the United States but approximately the same share of income. However, there are significant disparities across income groups in the ________________________________________________________ language are likely to be part of the “causal process” that influences the selection of center care, “which in turn implies the use of a subsidized slot.” 36 share of income going to pay for child care. Poor families who do pay for care (57 percent of families using child care) pay an average of 24 percent of earnings, compared to about 7 percent for families with income above the state median. Twenty-eight percent of families receive free care, either because they receive government or social services assistance (including free care through Head Start) or because relatives provide free care to the children. Higher-income families pay more in absolute terms than do lower-income families, but lower-income families pay twice as large a share of their income for child care. Most families at or below 75 percent of the SMI receive some assistance in paying for child care. Forty-six percent of poor families receive government or social services assistance in paying for care. Onethird of former welfare recipients and two-thirds of current recipients who use child care while working receive government assistance in paying for care. However, current welfare recipients are less likely to receive free care from relatives, perhaps because the welfare system will reimburse relatives for providing care. However, use of structured care is much higher among welfare recipients and other low-income families receiving assistance. Reducing eligibility for subsidies will increase the costs of work for many families, increasing the share of income spent on child care. Although relatives caring for young children for free may absorb some of the costs, reduced subsidies may also force some parents currently using structured care settings to switch to more informal arrangements. 37 5. Preschool Enrollment Child care subsidies were dramatically increased in the late 1990s as an underpinning for welfare reform, but current policy trends concentrate on the role of child care as early childhood education. The foundational mission of First 5 programs in California, programs for children ages 0–5 financed through a 50 cent per pack cigarette tax approved under Proposition 10, is built on the conclusion that “current research in brain development clearly indicates that the emotional, physical and intellectual environment that a child is exposed to in the early years of life has a profound impact on how the brain is organized [and] . . . significantly influence[s] how a child will function in school and later in life.” Indeed research on early childhood intervention programs such as those reviewed by Karoly et al. (1998) and Gomby et al. (1995) found that intensive early childhood education programs had positive effects on cognitive and school outcomes, as well as social outcomes, with significant long-term cost savings. This work feeds into larger state policy concerns about school outcomes and school readiness from intensive early childhood education programs. As a result, one policy recommendation that has been gaining ground is a call for universal preschool. Voluntary universal preschool for children ages 3 and 4 was recommended as part of California’s new Master Plan for Education, because “the state has a profound interest in making available . . . the early education opportunities that support a child’s emotional, social, physical, linguistic, and cognitive development” (California State Senate, 2002). The Packard Foundation has made universal preschool a priority area for future grantmaking, Los Angeles County’s First 5 commission has dedicated an initial $100 million toward universal preschool in that county, and other county commissions are conducting feasibility studies for similar projects. As a background for understanding the potential desirability of universal preschool, this chapter examines more closely care settings for 39 preschool-age children. The following chapter presents a framework for estimating the costs of such proposals. Care Settings for Preschool-Age Children The call for universal preschool has particular relevance for California, where preschool-age children—those ages 3–5—are less likely than children elsewhere to be enrolled in a structured, center setting. In the previous chapter, we considered all types of care selected, recognizing that children were commonly cared for in multiple settings. To understand the issue of universal preschool, we need to consider whether children receive any school or center care.1 For this reason, we divide preschool-age children into four exclusive categories of care: • Not in care: Children cared for by their parents only. • Kindergarten: Among children in any care arrangement, we first distinguish children attending kindergarten. We classify them as kindergartners, although they may also receive care in other settings, including centers. • Preschool: Children who do not attend kindergarten but who do receive another type of structured care (preschools, nursery schools, Head Start programs, or child care centers) are classified as attending preschool. These children may also receive care in other settings. • Other care: Children in care who do not attend any kindergarten or preschool programs are classified as enrolled in other care settings. These children may be in more than one non-kindergarten, non-preschool arrangement. As Figure 5.1 shows, 57 percent of children ages 3–5 in California are enrolled either in preschool or in kindergarten, compared to 63 percent in the rest of the nation. More California children are enrolled in kindergarten—19 percent compared to 13 percent elsewhere—because ____________ 1Developmentally rich care may occur in non-center settings, and not all center programs are developmentally rich. However, most universal preschool/prekindergarten initiatives rely on center programs as well as schools. 40 Percentage 60 50 *50 40 *38 30 20 10 0 Preschool California Rest of United States *19 *13 19 20 *24 *17 Kindergarten Other care Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.1—Child Care Settings for Children Ages 3–5 California admits children at a younger age than the national average.2 However, a significant gap in preschool attendance leads to a large California-U.S. gap in enrollment in early childhood education. Enrollment in preschool and kindergarten varies greatly by age. Figure 5.2a, 5.2b, and 5.2c break down care settings by age within the preschool group. In our data, only children age 5 are enrolled in kindergarten, including 54 percent of children that age in California. Among children age 5, in fact, early enrollment in kindergarten brings the total share in structured settings in California above the average of the rest of the United States, with 88 percent in preschool or kindergarten, compared to 85 percent in the rest of the country. For children ages 3 and 4, however, the story is very different. There is a 13 percentage-point gap in preschool enrollment of children age 4 between California children and other U.S. children. There is a 10 percent gap for children age 3. ____________ 2California schools must admit children into kindergarten at the beginning of the school year if they will be five years of age on or before December 2 of the school year. Legislation passed in 2000 allows for pilot programs requiring that children be age 5 by September 1, but no such pilot programs have yet been funded. 41 Percentage 70 60 50 *46 40 *34 30 20 10 0 Preschool *54 *39 California Rest of United States Kindergarten 8 3 Other care *9 *7 Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.2a—Child Care Settings for Children Age 5 70 *63 60 50 *50 40 30 20 10 0 Preschool California Rest of United States 22 19 *28 *18 Other care Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.2b—Child Care Settings for Children Age 4 Percentage 42 Percentage 70 60 50 40 *38 30 *28 20 10 0 Preschool 35 35 California Rest of United States *36 *27 Other care Not in care NOTE: Because many children are cared for in multiple settings during a typical week or month, the percentages across settings total more than 100. * Indicates significance at the 95 percent level. Figure 5.2c—Child Care Settings for Children Age 3 The difference in preschool enrollment is driven primarily by a larger share of California children not being placed in any non-parental care setting, rather than by greater use of other child care options, such as family day care, care by relatives, or care by a nanny or babysitter. For example, we find that California children age 3 were equally likely to be in other care settings but 9 percentage points more likely than other American children to be cared for only by their parents. Children not in any care setting also explain most of the difference for children age 4. Although children age 5 who are not yet in kindergarten would be a target group for universal preschool programs, it is easier to consider the issues for preschool attendance by focusing on children ages 3 and 4. For this reason, we will first look only at preschool attendance for these children, and then return to the issue of kindergarten later in the chapter. Characteristics Linked to Preschool Attendance The lower workforce participation of California parents partially explains the higher level of parental-only care in California, but preschool enrollment is lower in California regardless of employment status. Excluding children age 5, 53 percent of preschool-age children in 43 California have at least one parent not in the workforce, compared to 45 percent of children elsewhere. However, the gap in preschool enrollment is even larger for children ages 3 and 4 with both parents employed (or with an employed single parent)—47 percent are enrolled in preschool in California, compared to 60 percent elsewhere. The U.S.-California gap in preschool enrollment persists across every demographic group, including family structure, child’s ethnicity, mother’s education, and family income (Table 5.1). For virtually all of these demographic categories, California has both more children in groups less likely to participate in preschool and lower participation within the subgroup. For example, 26 percent of children ages 3 and 4 in California live in poverty, compared to 21 percent elsewhere. Throughout the United States, children from poor families are less likely to be enrolled in preschool, but the gap between U.S. and California preschool enrollments is larger for poor families than for any other income category. Fewer California children come from the high-income or high-education groups that were most likely to participate in preschool, but enrollment is lower in California even for these more advantaged groups. In looking at California only, we also find that regional differences as well as differences in the number of licensed child care slots in the county of residence are linked to differences in preschool enrollment. As Figure 5.3 shows, Los Angeles and the Farm Belt counties of the Central Valley and Central Coast have the lowest preschool enrollments, whereas the Bay Area and the North and Mountain counties have the highest enrollment.3 In part, this is linked to the licensed child care capacity in these regions, measured by the ratio of children to licensed child care slots (Figure 5.4).4 The Bay Area and the North and Mountain counties have the lowest ratio of children per licensed slot available, and Los ____________ 3See the appendix for a list of counties included in each region. 4This measure, developed by the California Child Care Resource and Referral Network, divides the children of working parents in the population (based on census counts) by the number of slots in centers and licensed family day care. The statewide average for this measure was 4.7 children (up to age 13) for each licensed slot. 44 Table 5.1 Percentage of Children Ages 3 and 4 in Preschool, by Demographic Group California Rest of United States Parental employment All parents employed Not all parents employed Parent works evenings/weekends Yes No Family structure Single parent Two parents Child’s race/ethnicitya Non-Hispanic white Hispanic Parental educationb Less than high school High school diploma/GED Some vocational education/college Associate degree or more Other adults in household No Yes Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare 47* 34 42* 49 52* 36 45 35* 30* 35 39* 55* 42 34* 29* 36* 48* 51 50* 33 40* 33* 60* 41 55* 62 57* 50 52 35* 31* 46 48* 64* 52 45* 42* 43* 52* 63 60* 40 47 53* NOTES: Italics indicate significance at the 95 percent level. All differences between California and rest of the United States were statistically significant. aSample sizes in California are too small to report breakouts for black and Asian children. bBased on years of education of the “most knowledgeable adult” for the focal child. *Indicates significant differences from reference group 45 Percentage 70 60 *57 50 *50 *43 40 *31 30 32 20 10 0 Bay Area Los Angeles Other Southern Farm Belt California North and Mountain * Indicates significance relative to the Farm Belt at the 95 percent level. Figure 5.3—Share of California Children Ages 3 and 4 in Preschool, by Region Percentage 70 60 Higher Capacity Lower *51 50 *50 40 30 *28 20 10 0 Fewer than 4 children per licensed slot 4–5 children per 5–6 children per licensed slot licensed slot 34 More than 6 children per licensed slot * Indicates significance relative to the lowest capacity group at the 95 percent level. Figure 5.4—Share of California Children Ages 3 and 4 in Preschool, by Child Care Capacity in County of Residence 46 Angeles has the highest ratio. Not surprisingly, the takeup of preschool is higher in counties with more child care slots available. However, we cannot distinguish between supply-driven differences (where enrollment is lower because capacity is lower) and demand-driven differences (where capacity is lower because fewer families are seeking preschool). The high availability of licensed child care in the North and Mountain counties is largely due to state-subsidized preschools in this region, along with relatively fewer other options. In this region, the high capacity reflects a large supply that may not be fully utilized. Because many of the characteristics in Table 5.1 occur together—for example, low-income families often have lower educational attainment— Table 5.2 provides regression results assessing the marginal contribution of different factors in explaining the likelihood that a child age 3 or 4 attends preschool in California.5 Some socioeconomic disadvantages reduce the chances of a child attending preschool. For example, children with less-educated mothers or from lower-income families are less likely to attend preschool. Income is one of the most important factors determining preschool attendance in California. Controlling for other factors, and for place of residence, the probability of attending preschool rose 4 percent for each additional $1,000 of family income. Compared to children whose mothers just completed high school or earned a GED, children whose mothers did not finish high school were 3 percent less likely to attend preschool, and those with more education were 3 to 4 percent more likely. Welfare recipients and low-income mothers who had never received welfare were 6 to 7 percent less likely than higher-income mothers to have their children enrolled in preschool. However, controlling for other factors, mothers who had previously received welfare were more likely than mothers with other welfare statuses to have their children in preschool— a result that is consistent with the observation, based on the use of child care vouchers, that families who receive child care assistance while on ____________ 5Consistent with Table 5.2, Appendix Table A.4 presents regression results for the United States with and without interaction terms for California. Appendix Table A.5 presents California regression results with and without regional variables. 47 Table 5.2 Regression Results on the Likelihood of California Children Ages 3 and 4 Being Enrolled in Preschool Household composition and employment Single parent, employed Married/partners, both employed full-time At least one parent not working full-time (married or single) Parent works evenings or weekends Family income For each additional $1,000 of annual family income Mother’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Mother’s welfare status Currently on welfare Formerly on welfare Never on welfare—income below 75% SMI Never on welfare—income above 75% SMI Other adults in home Hispanic child Child is age 3 Region of state Bay Area Los Angeles Other Southern California Farm Belt North and Mountain Children per licensed slot in the county % Change in Likelihood 36* 16* – 2* 4* – 3* 4* 3* – 6* 4* – 7* – 9* 5* – 16* 5* – 7* 8* – 4* – 5* NOTE: Italics indicate reference group. *Indicates significant differences at the 99 percent level. Marginal effects calculated from probit. welfare often shift to more structured care settings after they transition off welfare. Surprisingly, single parenthood does not appear to be a disadvantage in terms of preschool attendance. Compared to children of married parents who did not both work full-time, the children of single parents were 20 percent more likely to attend preschool. Obviously, married parents may choose to work part-time or not at all specifically to stay 48 home with young children. Yet even among unemployed single parents, the probability of their children attending preschool was dramatically higher. The explanation for this is not immediately clear. Without controlling for other factors, employed single parents are the most likely to have their children in preschool, followed by dual-income two-parent families. However, non-employed single parents are more likely than other non-working parents to have a child in both preschool or other care settings (Figure 5.5). Single parents may live in neighborhoods better served by preschools, perhaps because of greater population density in urban areas. Although single parents may also have greater access to free preschool through such programs as Head Start, low use of care by single parents in poverty suggests that such programs are not the primary drivers of this phenomenon. It is interesting to note that the much lower preschool participation for Hispanic children appears to be driven almost entirely by the economic and family characteristics of Hispanic families. Once we account for income, education, employment status, and other adults in the home, Hispanic children are actually more likely to attend preschool. Finally, the other effect that changes between the simple descriptive statistics and the regression findings is the preschool participation in the Percentage 80 70 65 69 52 50 46 40 38 30 21 20 Single parent, employed Single parent, not employed Two parents, both employed Two parents, one employed Two parents, not employed 73 44 30 30 23 17 31 18 10 0 Preschool 6 Other care 5 Not in care Figure 5.5—Child Care Settings for Children Ages 3 and 4, by Household Composition/Employment 49 North and Mountain counties. In Figure 5.3, we saw that this region had the second highest participation rate among the five California regions. However, once we control for other factors, especially the larger capacity of licensed slots in the region, the marginal role of this region of residence is negative; so all else being equal, children in this region are less likely to be enrolled in preschool. Targeting Efforts to Groups Less Likely to Use Preschool Efforts to improve preschool attendance in California may be most effective if they are tailored to address those populations least likely to use preschool. As we have seen, children ages 3–5 not in preschool fall into two categories: children in parental care only and children of working parents who are placed in other child care settings. In this section, we review the characteristics of each group to better understand the challenges in increasing preschool enrollment. Children in Parental Care Only Not surprisingly, the majority of children ages 3–5 who are not in any care setting are in two-parent families where one parent stays home, typically to serve as the primary caregiver. Preschool-age children are more likely to be cared for only by parents when there are other children in the household. Just 16 percent of only children are cared for exclusively by parents, compared to 28 percent of children with one sibling under age 13 and 38 percent of children with two or more siblings (Figure 5.6). This effect is particularly strong if the sibling is younger than age 3. At the same time, however, children cared for by parents only are more likely to come from disadvantaged households. Sixty-four percent of these children have mothers whose education ended at or before high school graduation, and a disproportionate share are Hispanic.6 Finally, these children are much more likely to be in low-income households. Poor children are 70 percent more likely to be cared for only by parents, ____________ 6Liang, Fuller, and Singer (2000) found that when controlling for economic and education attributes, the lower probability of Hispanic parents than whites using center care diminishes, as was the case with other adults in the household. 50 Percentage 80 70 60 57 50 50 40 38 30 20 10 0 Preschool One child under age 13 Two children Three or more children 27 22 24 38 28 16 Other care Not in care Figure 5.6—Child Care Settings for Children Ages 3–5, by Number of Children Under Age 13 largely at the expense of preschool attendance, and near-poor children are also less likely to be placed in a non-parental setting. In fact, for preschool attendees, the median household income is $41,700 whereas the child not placed in care comes from a household with a median income of $25,000. Assessing the characteristics of children not in child care arrangements through the prism of universal preschool, however, reveals another pattern that is important for school-readiness policies: Not only are these children less likely to attend preschool, but they are also less likely to attend kindergarten at age 5. Statewide, 54 percent of children age 5 attended kindergarten, compared to only 49 percent of low-income children (children from families below 50 percent of the state median income). A similar pattern holds for children of less educated mothers: 60 percent of children age 5 whose mothers had an associate degree or more education attended kindergarten, compared to only 49 percent of those whose mothers did not complete high school. Much of the difference is driven by differences in the Hispanic community. Only 43 percent of Hispanic children age 5 attended kindergarten, 16 percentage points lower than the share of non-Hispanics. (The lower use of kindergarten by Hispanic families holds outside California as well.) 51 Children in Other Care Settings Increasing preschool enrollment for children in other care settings is a somewhat different challenge. These families have chosen to place their children in a care setting, but whether because of the characteristics of the caregivers, the costs of care, or the convenience of the providers, they have selected settings other than preschool. In fact, 62 percent of these children are cared for by relatives, most commonly in the child’s home (Figure 5.7). Relatives may be the most affordable option for these families, but there may also be a strong preference for this type of care. Many of these relatives provide care in the child’s home. Working families are most likely to rely on care settings other than preschool. Use of other care settings is almost twice as common for children ages 3 and 4 with both parents working (or a working single parent) than for children with a parent at home: 38 percent versus 20 percent. Other characteristics associated with an increased likelihood of selecting care outside preschool settings include current welfare receipt, fewer children in the household, and lower income. An important factor in parents’ selection of settings other than preschool is the ability to have one consistent child care provider. More than half of children ages 3 and 4 in preschool regularly spend time in more than one care arrangement, compared to only 19 percent of children in other care settings. For families without a parent at home, 80 70 62 60 Percentage 50 40 36 30 20 12 10 0 Family day care Nanny/babysitter Care by relative Figure 5.7—Child Care Settings for Children Ages 3–5 Not in Preschool 52 the share of families who combine preschool with another setting rises from 52 percent to 58 percent. This may be because most preschool programs are not full-time, as shown in Table 5.3. Only 33 percent of children in preschool are in a program for more than 30 hours per week, and the majority are in programs for fewer than 20 hours per week. Half-time programs are particularly common for children with a parent at home, whereas full-time programs are more common for families with both parents in the workforce. Half the children in full-time preschool programs are still in other care settings, predominantly children without a parent at home. In light of the complexity of arrangements required when combining preschool with other care arrangements to accommodate work schedules, it is not surprising that many parents choose other care settings over preschool. Table 5.3 Share of Preschool Arrangements That Are Part-, Moderate-, and Full-Time Hours in Preschool Part-time (<20 hr/wk) Moderate time (20–29 hr/wk) Full-time (30+ hr/wk) % of Children Ages 3 and 4 in Preschool 52 15 33 % by Parents’ Employment Status Parent(s) Parent at Employed Home 43 64 16 15 48 20 NOTE: Children may be in more than one preschool arrangement. Promoting Use of Preschool California families are significantly less likely than families in the rest of the nation to enroll children ages 3–5 in preschool programs. Given that the Master Plan for California Education has identified rising concerns about school readiness for children entering kindergarten, our analysis of the characteristics of families of these children in California, their choice of care, and the costs of care points to some possibilities for increasing preschool enrollment in California. Although we do not have enough information to determine how reducing the costs of preschool would affect parents’ choice of settings, the findings in our study do 53 suggest several recommendations for encouraging enrollment in preschool: • Encourage kindergarten enrollment. Although California’s relatively early age cutoffs for kindergarten result in higher enrollment in California than elsewhere in the United States, kindergarten enrollment for children age 5 is disproportionately low among low-income families, children with less-educated parents, and especially among Hispanic families. • Focus on mothers with lower educational attainment. More than 83 percent of preschool-age children not attending preschool or kindergarten are cared for by their parents or other relatives. Some of these families may face economic barriers to enrolling their children in preschool, but controlling for income, mothers with lower educational attainment are less likely to place children in preschool. Targeted efforts to promote the use of preschool may be appropriate for some of these families, especially those who may not take advantage of opportunities offered by Head Start and other currently funded programs.7 • Offer more full-day preschool programs. State preschools in California traditionally offer half-day programs, which may not be attractive to working families. If this model is followed for universal preschool, families with both parents working full-time or poor single-parent families may continue to rely on care settings that better accommodate work schedules. Full-day fullyear programs or programs that help parents with multiple providers may be more attractive, although this strategy could be substantially more expensive. ____________ 7This recommendation assumes that programs such as Head Start and state contracted centers are available and accessible. 54 6. Costs of Universal Preschool The cost of care for universal preschool depends on a number of factors. First, there are the factors that determine how many children will enroll in preschool funded through a universal program, including eligibility rules and the share of eligible children who choose to enroll. A related issue is the share of children currently enrolled in privately paid child care who move into the newly funded program. Finally, there are the determinants of the costs per child. In the remainder of this chapter, we describe alternative assumptions that could be made for each of these factors and then explore the range of costs implied by the different scenarios. Assumptions on Takeup of and Costs for Preschool The first task in estimating the costs of universal preschool is to estimate the number of children likely to enroll. To determine this, we must first consider who would be eligible for universal preschool. Although “universal” suggests that all children would be eligible, the initial drive could target only low-income children. The Master Plan and proposed legislation suggest that children ages 3 and 4 would qualify. Presumably, children age 5 who do not meet the cutoff for kindergarten would also qualify. The second consideration is what share of parents will choose to enroll their eligible children, since many parents may continue to prefer their current care arrangements, whether parental care or in other informal settings. Of course, some of the children who would qualify for free care under the universal preschool plan are already enrolled in preschool. Takeup of the universal program by these children will increase program costs without increasing the number of children actually enrolled in preschool. How many will move into this program from existing preschool providers depends in part on which providers are included in the program. Some models of universal preschool rely on schools to offer the care; others permit existing child care centers to 55 qualify as providers. Typically, universal preschool programs have standards for providers that are more rigorous than licensing requirements. If providers can easily convert to a universal preschool program, then the children in their care would presumably shift from private payment and voucher assistance to the free program. If providers cannot easily convert, parents whose children are eligible for the universal program might still keep their children with these providers, because they prefer the provider (because of location, convenience, appropriateness for their child, or other reasons) or because they cannot access the free care. Finally, many preschool children already receive free care through Head Start, state preschools, state contracted slots, or subsidized care. We do not count these children as imposing new costs on the system, although they may shift between child care settings. We have considered different scenarios for each of these factors. In addition, we have considered different strategies for calculating the average costs per child in the program. Families Eligible for Universal Preschool We have examined four eligibility criteria for universal preschool. The broadest definition is “universal,” meaning all children of the appropriate age would be eligible to receive free preschool. At its broadest, this definition would include all children ages 3–5 in California, except those age 5 enrolled in kindergarten. A more constrained version would limit eligibility to those families currently eligible for child care subsidies—a definition that would include only children from families with income below 75 percent of the state median income. However, because universal preschool is focused on child development, children would be eligible whether or not their parents are in the workforce, a difference from the subsidy system. Our third option is a variation on this, limiting eligibility to children with family income below 50 percent of the SMI. Current subsidy policy does not require co-payments for families in this income range. Finally, we consider narrowing eligibility to children with family income below poverty. These children are already eligible for preschool through the Head Start program but slots are not necessarily available to serve all children. 56 Enrollment of Children Not Currently in Preschool One would never expect all eligible children to actually enroll in a voluntary universal preschool program. Short of 100 percent enrollment, we have developed two alternative scenarios for the takeup of universal preschool. Our high estimate assumes that preschool enrollment for participating children matches that of wealthy families. To make this truly an upper bound, we have chosen a higher income threshold than the median income, which defined our top income category above. For this exercise, we have selected an income threshold of 150 percent of the SMI to define a group of children whose parents faced much lower economic barriers to preschool enrollment. We are assuming, essentially, that in the absence of economic barriers to preschool enrollment, the underlying preferences for preschool over other care options are identical between high- and low-income families.1 Our low estimate was set arbitrarily at half of our high estimate. To calculate enrollment based on the high estimate assumption, we have compared enrollment of California families qualifying under each eligibility scenario to enrollment for children of the same age in families with income above 150 percent of the state median income. Crowd-Out of Existing Preschool If universal preschool were available, many if not most children currently enrolled in preschool would switch to the universal preschool program. For our high estimate, we assume that all children enrolled in preschool paid for by their parents would move from privately paid preschool to the new program. Children already enrolled in free preschool programs, such as Head Start, are not included in our count of new enrollment costs. Existing free preschool care may become part of the universal program, but this would not incur new costs. Therefore, our high estimate uses the NSAF data to determine the share of families (in different eligibility groups) enrolled in preschool care with any outof-pocket child care expenses. This is probably an upper bound for two ____________ 1Realistically, the cost of preschool is still perceived as a potential barrier to enrollment for higher-income families. Nor do we have a good understanding of what different families’ preferences would be in the absence of any cost for preschool. 57 reasons. First, because of small sample sizes in the various eligibility groups, we use U.S. estimates of the share of children in paid preschool care, which is higher than the share paying for care on average in California. Second, because out-of-pocket costs are recorded for all care settings and all children, it is possible that we are overcounting the share of families paying for preschool when their children are in multiple settings. Some private paying parents may in fact choose to keep paying for preschool even when their children are eligible for the new program. How many will depend on the way the universal preschool is implemented. For example, many more would switch if most existing preschool providers are able to participate in the universal program. Fewer would switch if the universal care is only half-day or provided only in limited locations. We assume that higher-income families are less likely to switch than lower-income families. Thus, our low estimate starts with the share paying for preschool in California. We then assume that 50 percent of families with income above 75 percent of the SMI will switch to the new program, as will 70 percent of families between the poverty threshold and 75 percent of SMI and 90 percent of poor families. Using these assumptions, Table 6.1 shows a variety of scenarios for takeup of universal preschool by eligibility criteria and age. It also shows our two assumptions for the share of children currently in preschool who would also move to free care. Because poorer families are the least likely to participate in child care currently, a larger share of children would need to move from parental or other care into preschool to match the enrollment of wealthier families. The takeup estimates are fairly similar for the two low-income eligibility standards—around 60 percent. For all families to match the preschool use of wealthy families, 53 percent of children not currently in preschool would have to move into such arrangements. The higher share for children age 5 reflects the lower use of both preschool and kindergarten by lower-income families. Some of these children would be eligible for kindergarten. However, we assume that a somewhat larger share enter preschool instead, assuming that parents of these children feel that their children are not ready for school. 58 Table 6.1 Assumptions on Takeup of Universal Preschool for California Children All families eligible All preschoolers Age 3 Age 4 Age 5 Eligible if income below 75% SMI All preschoolers Age 3 Age 4 Age 5 Eligible if income below poverty All preschoolers Age 3 Age 4 Age 5 aExcludes kindergartners. Enrollment as % of Children in Care Other Than Preschoola High Low Estimate Estimate 53 27 51 25 53 27 62 31 58 29 57 29 57 28 68 34 60 30 59 29 57 29 68 34 % of Children in Preschool Moving to Free Care High Low Estimate Estimate 62 37 50 26 38 25 Among families already using preschool, low-income families are much more likely to have free care. For this reason, the more constrained the eligibility standards, the smaller the share of children moving into the new program. Thus, more generous eligibility standards increase not only the total number of children who could switch from family-paid preschool to state-funded preschool but also the share of those we would expect to switch. Cost per Child in Universal Preschool Policymakers would have more control over the cost per child for universal preschool, although they would still face constraints. In particular, the cost per child would have to be sufficient to encourage the supply expansion required to absorb all the additional children served. In addition, the costs would need to cover care of a sufficiently high quality to justify choosing preschool over other care arrangements. There are three potential models for the cost of care: market prices, state 59 preschool allocations, or the standard reimbursement rate (SRR) provided for children in contracted slots in child care centers. State preschool is a part-day, part-year program, whereas the contracted slots are usually full-day, full-year.2 Similarly, parents in the broader child care market use both full-time and part-time care. Alternative Estimates of the Costs of Universal Preschool Using the set of options described above, Tables 6.2 and 6.3 show the number of children served and the costs of care in our various scenarios based on the number of children ages 3–5 in the 2000 Census. We consider five options for the price of care per child. The first three are market-based prices. The NSAF estimate of $4,216 annually Table 6.2 Estimates of Average Annual Costs for Universal Preschool: High Estimate ($ millions) All families eligible All preschoolers Age 3 Age 4 Age 5 Eligible if income below 75% SMI All preschoolers Age 3 Age 4 Age 5 Eligible if income below poverty All preschoolers Age 3 Age 4 Age 5 No. of Children Served 726,402 272,626 301,308 152,468 404,493 169,524 161,701 73,268 205,163 76,790 87,211 41,162 NSAF $4,216/ yr 3,062 1,149 1,270 643 1,705 715 682 309 865 324 368 174 RMR Survey Part-Time Full-Time $3,200/yr $5,800/yr 2,324 872 964 488 4,213 1,581 1,748 884 1,294 542 517 234 2,346 983 938 425 657 1,190 246 445 279 506 132 239 State Preschool 2,348 881 974 493 1,308 548 523 237 663 248 282 133 SRR 4,881 1,832 2,025 1,025 2,718 1,139 1,087 492 1,379 516 586 277 ____________ 2For simplicity, we do not adjust assumptions on the likelihood of switching from privately paid preschool to the universal program to reflect whether the proposed program is part-day, part-year or full-day, full-year. 60 Table 6.3 Estimates of Average Annual Costs for Universal Preschool: Low Estimate ($ millions) All families eligible All preschoolers Age 3 Age 4 Age 5 Eligible if income below 75% SMI All preschoolers Age 3 Age 4 Age 5 Eligible if income below poverty All preschoolers Age 3 Age 4 Age 5 No. of Children Served 401,335 145,673 167,649 88,013 206,242 85,567 82,930 37,745 110,825 39,621 48,316 22,888 NSAF $4,216/ yr 1,692 614 707 371 869 361 350 159 467 167 204 96 RMR Survey Part-Time Full-Time $3,200/yr $5,800/yr 1,284 466 536 282 2,328 845 972 510 660 1,196 274 496 265 481 121 219 355 643 127 230 155 280 73 132 State Preschool 1,298 471 542 284 667 277 268 122 358 128 156 74 SRR 2,697 979 1,127 591 1,386 575 557 254 745 266 325 154 for a child in preschool averages across children in part-time and fulltime settings.3 Two other estimates of the market price come from the regional market rate (RMR) survey conducted annually through 2001 to determine price ceilings for subsidy rates. The RMR survey collects data on child care fees for centers in every county in California, taking a census in small counties and a sample in larger counties. The survey reports fees separately for full-time and part-time care. These two options are shown as different scenarios for the price estimates. The part-time preschool rate from the RMR survey is nearly identical to the part-day, part-year rate provided through state preschools. The final cost option assumes the full-year, full-day SRR for state contracted slots. ____________ 3Because the NSAF reports costs per family across all arrangements, we use the U.S. estimates for the cost of care for children without siblings in preschool as the only setting. (Taking children in one setting only biases this price toward part-day programs.) The U.S. estimate is then inflated to account for average price differences between California and the rest of the United States. 61 The exact method for selecting the cost of care matters less than the distinction between full-time and part-time care: The estimates based on the SRR are very close to those based on the full-time RMR, and the estimates based on state preschool funding are very close to those based on the part-time RMR. Because the NSAF captures a blend of full-time and part-time care, the estimates based on the NSAF fall between the other pairs. A universal preschool program serving all preschool children fulltime, full-year could be expected to cost as much as $4.88 billion annually, assuming high takeup rates (Table 6.2). The lower estimates of takeup drop this to $2.70 billion (Table 6.3). As the eligibility criteria narrow—moving the program away from universal coverage and closer to a poverty program—the costs fall. A program supporting full-year, fulltime preschool care for children in poverty would cost between $1.38 billion (Table 6.2) and $745 million (Table 6.3). Part-day, part-year programs such as the current state preschools would cost about half as much as the full-time programs, with the costs at the broadest eligibility falling to between $2.35 billion and $1.3 billion. However, our estimates do not adjust for lower takeup of parttime care, given the reduced convenience for working families. Programs that permitted parents to choose between full-time and part-time care would fall between these cost estimates and would also have the highest expected enrollment rates. A substantial portion of these costs would go toward serving families who are already enrolling their children in preschool. For the three eligibility options, Table 6.4 shows the share of costs resulting from children currently enrolled in preschool. The share of costs from currently enrolled children falls as eligibility is constrained, both because poorer children are more likely to receive free preschool in the absence of a universal preschool program and because total enrollment in preschool rises with income. Below the 75 percent of SMI threshold, the different assumptions on takeup by children already enrolled in preschool have relatively little effect. If all families are eligible, this share ranges from 50 to 54 percent; for children in families with income below 75 percent of SMI, the share of costs from currently enrolled preschoolers falls to about 34 percent. In fact, the high and low estimate assumptions matter 62 Table 6.4 Percentage of Costs Resulting from Participation of Currently Enrolled Preschoolers, by Takeup Assumptions Eligibility Criteria All families eligible Income below 75% SMI Income below poverty High Estimate 50 33 26 Low Estimate 54 34 32 somewhat more for the takeup by children not currently enrolled in preschool, so the share of costs resulting from currently enrolled children is slightly higher in the low estimate version. Finally, it is important to recognize that these cost assumptions abstract from the issue of available supply. In fact, it may be necessary to increase reimbursement rates to induce a sufficient increase in supply of preschools to accommodate children served under this program. Table 6.5 reports the additional slots that would be needed in our high and low takeup scenarios and for each eligibility criterion. These counts include children currently in parental care only or in settings other than preschool and kindergarten. Obviously, the number of additional slots needed is higher the broader the eligibility criteria and the higher the takeup by children not currently enrolled. (In our method, these factors are not affected by the choice of reimbursement rate.) The extra slots required range from around 76,000 in the lowest estimates to 366,000 in the highest estimates. This massive addition of slots could not be accommodated quickly by existing providers. Table 6.5 Additional Slots Needed for Children Enrolling in Preschool, by Takeup Assumptions Eligibility Criteria All families eligible Income below 75% SMI Income below poverty High Estimate 365,895 270,774 151,606 Low Estimate 182,947 135,387 75,803 63 Summary The costs of universal preschool vary widely depending on assumptions of eligibility criteria and takeup of preschool by children not currently enrolled. At the high end, full-year, full-day universal preschool could cost as much as $5 billion annually when fully implemented, although eligibility criteria more closely matched to the current subsidy system would reduce this cost to between $1.4 billion and $2.7 billion. Of course, the broader the criteria, the less the program may be perceived as a public assistance program and presumably the higher its political popularity. Similarly, if universal preschool were modeled after state preschools, offering part-day, part-year programs, it would be significantly less expensive. The method for estimating the average cost of care matters less than the distinction between full-time and part-time programs. Parttime programs are much less accessible to working parents. Given the significant gap in preschool use between high-income and low-income families with a parent not in the workforce, this limitation may be a secondary concern. Finally, it is critical to note that these costs do not account for the administrative and other infrastructure costs that would be required to run a universal preschool program. In particular, if large numbers of children do move into preschool, there may be substantial startup costs to ensure an adequate number of facilities and classrooms to accommodate the new enrollment. 64 7. Conclusions As state resources become ever more limited, it is difficult to maintain the existing balance between child care policy intended to support working parents and policy to promote developmentally enriched care. At present, child care subsidies provide an important resource to parents who need child care to permit them to work. These subsidies have been a particularly important resource for families transitioning from welfare to work. These families, as with other lowincome families, spend a disproportionate share of their income on child care. Yet many families who are eligible for child care do not receive government assistance because the pool of subsidy funds is not sufficient to cover all eligible children. Without additional funding, the only way to support more children is to provide less support to those currently covered. Since assistance seems to encourage families to use more formal, presumably more enriched care, and such options as increased family fees would provide disincentives for families choosing higher-priced care, this tradeoff could also discourage assisted families from using high-quality care. At the same time, such proposals as the call for universal preschool are seeking to substantially increase spending on developmentally rich early childhood education programs. Although the existing subsidy system offers substantial parental choice, universal preschool would provide very generous benefits for enrolling in high-quality programs. The hope is that increased school readiness preparation, especially for disadvantaged children, would pay off down the road in improved school outcomes. Advocates of universal preschool view the link to school readiness and universal coverage as important selling points for these proposals. Nevertheless, universal preschool may cost several billion dollars annually, with as much as 50 percent of the cost offset by private dollars spent by families with children currently enrolled in preschool. 65 The tradeoff between these two goals of work support and developmental care is a political choice. Whatever is chosen, it is important to address the needs of disadvantaged families, which are paying 20 percent or more of their earnings toward child care with relatively low access to the more formal care sectors. In addition, the low participation of children of parents with low educational attainment and of Hispanic children in preschool or other care settings suggests that to improve school readiness for these populations, additional efforts may be needed to increase awareness of the advantages of formal programs of early childhood education. 66 Appendix Supplementary Tables Table A.1 Sample Characteristics for Family-Level Data (Costs) Most knowledgeable adult’s race/ ethnicity Non-Hispanic white Hispanic Other Income as a % of federal poverty threshold <100 100–199 200+ Welfare history Currently on welfare Formerly on welfare Never on welfare Most knowledgeable adult’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Number of children under age 13 1 2 or more California Weighted No. % Rest of United States Weighted No. % 483 41 14,004 69 599 40 2,856 11 201 19 3,738 20 364 26 4,394 20 344 23 5,335 23 575 51 10,599 57 154 11 1,220 6 156 10 3,048 13 966 79 15,926 81 264 21 2,211 11 357 25 6,351 32 292 23 4,705 23 348 31 6,842 33 402 40 7,011 44 881 60 13,279 56 67 Table A.2 Percentage of Children Ages 0–5 in Rest of United States in Child Care, by Demographic Group and Type of Arrangement % in NonParental Care % in Structured Care Care by Relative in Child’s Home % in Arrangementa Care by Relative Outside Child’s Family Home Day Care Nanny/ Babysitter 68 Family structure Single parent Two parents Child’s race/ethnicityb Non-Hispanic white Hispanic Parental educationc Less than high school High school diploma/GED Some vocational education/college Associate degree or more Other adults in household No Yes Income category Below poverty Below 75% of SMI 75–100% of SMI Above SMI 81 70 74 63 57 71 74 79 74 73 64 69 74 81 58 57 57 48 49 53 56 63 59 48 56 51 57 62 32 23 24 31 32 28 26 21 21 50 30 28 24 21 29 20 7 26 21 14 27 22 15 30 14 9 30 11 5 31 21 7 28 23 10 22 21 19 28 21 13 21 16 4 30 15 8 31 20 8 27 22 10 22 23 17 Table A.2 (continued) 69 Welfare receipt Never on welfare—income above 75% SMI Never on welfare—income below 75% SMI Formerly on welfare Currently on welfare % in NonParental Care 79 65 75 62 % in Structured Care 60 49 61 65 Care by Relative in Child’s Home 22 29 30 28 aChildren may be in more than one arrangement. bSample sizes in California are too small to report breakouts for black and Asian children. cBased on years of education of the “most knowledgeable adult” for the focal child. % in Arrangementa Care by Relative Outside Child’s Family Home Day Care 23 22 31 19 30 18 26 13 Nanny/ Babysitter 16 9 7 6 Table A.3 Counties Included in Each Region Other Southern Bay Area Los Angeles California Farm Belt Alameda Los Angeles Orange Colusa Stanislaus Contra Costa Riverside El Dorado Sutter Marin San Bernardino Fresno Tulare Napa San Diego Glenn Yolo San Francisco Santa Barbara Imperial Yuba San Mateo Ventura Kern Santa Clara Kings Santa Cruz Madera Solano Merced Sonoma Monterey Placer Sacramento San Benito San Joaquin San Luis Obispo North and Mountain Alpine Plumas Amador Shasta Butte Sierra Calaveras Siskiyou Del Norte Tehama Humboldt Trinity Inyo Tuolumne Lake Lassen Mariposa Mendocino Modoc Mono Nevada 70 Table A.4 Regression Results on Likelihood of Preschool Children Ages 3 and 4 Being Enrolled in California and Rest of United States Household composition and employment Single, employed Married/partners, both employed full-time At least one parent not working fulltime (married or single) Parent works evenings or weekends Income is $1,000 higher Mother’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Mother’s welfare status Currently on welfare Formerly on welfare Never on welfare—income below 75% SMI Never on welfare—income above 75% SMI Other adults in home Hispanic child Child is age 3 California resident % Change in Likelihood of Child Being in Preschool Rest of United Rest of States (No United States California Interaction) Interactions Interactions 12* 27* 18* – 1* 11* 7* – 5* 1* – 8* 3* 12* 3* – 0.5* – 11* – 6* 1* – 8* 3* 13* 3* – 2* – 12* 4* 3* 6* 1* – 8* – 8* 6* 5* – 9* – 9* – 22* – 10* – 9* – 14* – 23* – 49* 16* 2* 8* –– NOTE: Italics indicate reference group. *Indicates statistically significant at the 99 percent level. Marginal effects are calculated from probit. 71 Table A.5 Regression Results on Likelihood of Children Ages 3 and 4 Being Enrolled in Preschool in California Household composition and employment Single, employed Married/partners, both employed full-time At least one parent not working full-time (married or single) Parent works evenings or weekends Income is $1,000 higher Mother’s education Less than high school High school diploma/GED Some vocational education/college Associate degree or more Mother’s welfare status Currently on welfare Previously on welfare Never on welfare—income below 75% SMI Never on welfare—income above 75% SMI Other adults in home Hispanic child Child is 3 Region of state Bay Area Los Angeles Other Southern California Farm Belt North and Mountain Children per licensed slot in the county % Change in Likelihood of Child Being in Preschool No With Regional With Regions Variables Regions and Slots 40* 38* 36* 20* 18* 16* – 1* 4* – 2* 3* 6* – 4* 4* – 7* – 13* 4* – 15* – 2* 4* – 4* 4* 3* – 7* 4* – 7* – 9* 6* – 15* 12* – 11* 4* 0.5* – 2* 4* – 3* 4* 3* – 6* 4* – 7* – 9* 5* – 16* 5* – 7* 8* – 4* – 5* NOTE: Italics indicate reference group. *Indicates statistically significant at the 99 percent level. Marginal effects are calculated from probit. 72 Bibliography Brick, J. Michael, Pam Broene, David Ferraro, Tom Hankins, and Teresa Strickler, 1999 NSAF Sample Estimation Survey Weights, Urban Institute, Washington, D.C., July 2000, available at www.urban.org/ UploadedPDF/1999_Methodology_3.pdf. Brick, J. Michael, Gary Shapiro, Ismael Flores-Cervantes, David Ferraro, and Teresa Strickler, 1997 NSAF Snapshot Survey Weights, Urban Institute, Washington, D.C., July 2000, available at www.urban.org/ UploadedPDF/Methodology_3.pdf. California State Senate, Joint Committee to Develop a Master Plan for Education, The California Master Plan for Education, 2002, available at www.sen.ca.gov/masterplan. Cappiziano, Jeffrey, Sarah Adelman, and Matthew Stagner, What Happens When the School Year Is Over? The Use and Cost of Child Care for School-Age Children During the Summer Months, Urban Institute, Washington, D.C., June 2002, available at www.urban.org/ UploadedPDF/310497_OP58.pdf. Giannarelli, Linda, Sarah Adelman, and Stephanie Schmidt, Getting Help with Child Care Expenses, Urban Institute, Washington, D.C., February 2003, available at www.urban.org/UploadedPDF/310615_ OP62.pdf. Gomby, Deanna S., Mary B. Larner, Carol S. Stevenson, Eugene M. Lewit, and Richard E. Behrman, “Long-Term Outcomes of Early Childhood Programs: Analysis and Recommendations,” The Future of Children, Vol. 5, No. 3, Winter 1995. Hirshberg, Diane, Danny Shih-Cheng Huang, and Bruce Fuller, Welfare to Work and Child Care Selection: Which Families Use Subsidies and Home-Based or Center Care? Policy Analysis for California Education, 73 University of California, Berkeley and Stanford University, Berkeley, California, November 2002. Karoly, Lynn A., Peter W. Greenwood, Susan S. Everingham, Jill Houbé, M. Rebecca Kilburn, C. Peter Rydell, Matthew Sanders, and James Chiesa, Investing in Our Children: What We Know and Don’t Know About the Costs and Benefits of Early Childhood Interventions, RAND, Santa Monica, California, 1998, available at www.rand.org/ publications/MR/MR898/. Liang, Xiaoyan, Bruce Fuller, and Judith D. Singer, “Ethnic Differences in Child Care Selection: The Influence of Family Structure, Parental Practices, and Home Language,” Early Child Research Quarterly, Vol. 15, No. 3, 2000, pp. 357–384, available at www-gse.berkeley.edu/ research/PACE/ethnic_differences.pdf. Marrufo, Grecia, Margaret O’Brien-Strain, and Helen Oliver, Child Care Price Dynamics in California, Public Policy Institute of California, San Francisco, California, 2003. Sonenstein, Freya, Gary Gates, Stephanie Schmidt, and Natalya Bolshun, Primary Child Care Arrangement of Employed Parents, Urban Institute, Washington, D.C., May 2002, available at www.urban.org/ UploadedPDF/310487_OP59.pdf. 74 About the Authors MARGARET O’BRIEN-STRAIN Margaret O’Brien-Strain is an adjunct fellow at PPIC and Director of Social Policy at The SPHERE Institute. She is interested in a broad range of social services issues and has written about child care, home visitation programs, performance measurement, and welfare reform in California. She holds a B.A. in economics from Swarthmore College and a Ph.D. in economics from Stanford. LAURA MOYE Laura Moyé, formerly a research associate at PPIC, is interested in social issues related to child and youth well-being. She currently works in local government as a research analyst on child and youth development policy and programming. She holds a B.A. in urban studies from San Francisco State University and is currently pursuing an M.P.A. with a focus in policy analysis. FREYA LUND SONENSTEIN Freya Lund Sonenstein is Director of the Center for Adolescent Health at the Johns Hopkins University Bloomberg School of Public Health. She formerly directed the Population Studies Center at the Urban Institute in Washington, D.C., and was a visiting fellow at the Public Policy Institute of California in 2001. She is a Wellesley College graduate and holds a Ph.D. in sociology from Boston University. 75 Related PPIC Publications The Well-Being of California’s Children Frank F. Furstenberg, Maureen R. Waller, and Hongyu Wang What Happens to Families When They Leave Welfare? Thomas MaCurdy, Grecia Marrufo, and Margaret O’Brien-Strain Child Care Price Dynamics in California Grecia Marrufo, Margaret O’Brien-Strain, and Helen Oliver Expensive Children in Poor Families: The Intersection of Childhood Disabilities and Welfare Marcia K. Meyers, Henry E. Brady, and Eva Y. Seto “California’s Young Children: Demographic, Social and Economic Conditions” California Counts: Population Trends and Profiles Volume 4, Number 2, November 2002 Deborah Reed and Amanda Bailey A Portrait of Race and Ethnicity in California: An Assessment of Social and Economic Well-Being Belinda I. Reyes (editor), Jennifer Cheng, Elliot Currie, Daniel Frakes, Hans P. Johnson, Elizabeth Bronwen Macro, Deborah Reed, Belinda I. Reyes, José Signoret, and Joanne Spetz (contributors) Child Support and Low-Income Families: Perceptions, Practices, and Policy Maureen Waller and Robert Plotnick PPIC publications may be ordered by phone or from our website (800) 232-5343 [mainland U.S.] (415) 291-4400 [Canada, Hawaii, overseas] www.ppic.org 77" ["post_date_gmt"]=> string(19) "2017-05-20 09:36:11" ["comment_status"]=> string(4) "open" ["ping_status"]=> string(6) "closed" ["post_password"]=> string(0) "" ["post_name"]=> string(9) "r_1203mor" ["to_ping"]=> string(0) "" ["pinged"]=> string(0) "" ["post_modified"]=> string(19) "2017-05-20 02:36:11" ["post_modified_gmt"]=> string(19) "2017-05-20 09:36:11" ["post_content_filtered"]=> string(0) "" ["guid"]=> string(51) "http://148.62.4.17/wp-content/uploads/R_1203MOR.pdf" ["menu_order"]=> int(0) ["post_mime_type"]=> string(15) "application/pdf" ["comment_count"]=> string(1) "0" ["filter"]=> string(3) "raw" ["status"]=> string(7) "inherit" ["attachment_authors"]=> bool(false) }