oversight

Early Childhood Education: What Are the Costs of High-Quality Programs?

Published by the Government Accountability Office on 1990-01-24.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

Briefing Report to the Chairman,
Committee on Labor and Human



EARLY CHILDHOOD
EDUCATION
VVhat Are the Costs
of High-Qua&y
ROgUiJN!3?
      United States
GAO   General Accounting  Office
      Washington, D.C. 20548

      Human Resources    Division

      B-236126

      January 24,199O

      The Honorable Edward M. Kennedy
      Chairman, Committee on Labor
        and Human Resources
      United States Senate

      Dear Mr. Chairman:

      To assist the Congress in its deliberations on your proposed “Smart
      Start” bill (S. 123) and other pending legislation relating to early child-
      hood education and child care, you requested that we determine the
      costs of providing high-quality early childhood education. In later dis-
      cussions with your office, we agreed to (1) estimate the. average annual
      cost per child of providing high-quality early childhood education,
      (2) compare the average annual salary for early childhood education
      teachers with that of public elementary school teachers, and (3) deter-
      mine the extent to which the costs of a typical early childhood education
      center change when certain factors, such as the number of children
      enrolled and the ratio of teaching staff to children, change. This report
      elaborates on briefings provided to your office on May 25 and
      December 5, 1989.

      As currently proposed, S. 123 would provide financial assistance to
      state and local governments for early childhood education programs for
      prekindergarten-aged children (primarily 4-year-olds). Smart Start pro-
      grams would operate full day and full year; they would be required to
      meet certain criteria (see app. I). At least 67 percent of the funds allot-
      ted to the states under the bill would be targeted to centers to serve
      children from low-income families.

      Interest in expanding high-quality early childhood education programs
      for children from low-income families has grown as a result of both
      recent demographic trends and studies showing significant benefits
      from preschool programs. Of women with children under the age of 6,
      the percentage who are employed has tripled in the last three decades,
      from 20 percent to 63 percent; this trend is expected to continue. Thus,
      increasingly, children are being cared for by people other than their par-
      ents. Low-income children are much less likely to attend early childhood
      education programs than high-income children (33 percent versus 67
      percent). Yet, research has shown that high-quality early childhood edu-
      cation programs help low-income children through higher educational
      attainment and higher levels of employment later in life (see pp. 15-16).



      Page 1                                 GAO/HlUMO43BR   Early Childhood   Education
                                         B-236126




                                         To collect information on the costs of and services provided by high-
                                         quality early childhood education programs, we sent copies of a ques-
                                         tionnaire to the directors of 265 full-day, full-year preschool and early
                                         childhood education programs accredited by the National Association
                                         for the Education of Young Children (NAEYC). NAEYC is the only national
                                         accreditation system for early childhood education, and its criteria are
                                         similar to those in S. 123. For the most part, the centers surveyed had
                                         essentially the same program requirements as those included in S. 123.
                                         Most centers exceeded the bill’s requirements in some ways (child-staff
                                         ratio, daily group size, meals served, and teachers’ and teacher aides’
                                         training) (see p, 40).

                                         We believe the information obtained from the centers we surveyed gives
                                         a reasonable understanding of the costs of high-quality early childhood
                                         education in the United States; the data, however, were not meant to be
                                         nor are they necessarily representative of costs of all early childhood
                                         programs nationwide (see app. 11).Further, they are not necessarily rep-
                                         resentative of the costs of the programs reporting long-term benefits.


                                         In fiscal year 1988, on average, full-day, full-year NAEYC centers’ out-of-
What Are the Overall                     pocket costs, for all ages of children, were $4,200 per child; in-kind
Costs of the Centers                     donations for rent or mortgage, repairs, equipment and materials, and
We Surveyed?                             other items were estimated at an additional $600 per child. Centers were
                                         funded mostly through parent fees. Centers typically served about 80
                                         children. As shown in table 1, out-of-pocket costs varied by region of the
                                         country from about $3,900 in the West to about $4,900 in the Northeast.

Table 1: Average Annual Cost Per Child
Reported by NAEYC Centers                Average
(Fiscal Year 1988)                       costs per                                         Regions
                                         child               Nation   Northeast        Midwest           South             West
                                         Costs reported
                                         by NAEYC
                                         centers             $4,200       $4,949         $4,286          $3,978           $3,864
                                         Estimated
                                         costs if no in-
                                         kind donations
                                         were received        4,797        5,608          4,751           4,689            4,472




What Are the Major Center                 Personnel costs make up 65 percent of total costs for the centers sur-
Costs?                                    veyed. The other major center costs are rent and mortgage (11 percent)
                                          and miscellaneous operating expenses, such as educational materials



                                          Page 2                                   GAO/HRDBO-43BR    Early Childhood   Education
                          B-236126




                          and equipment, insurance, office supplies, repairs and maintenance, util-
                          ities, health and social services, and food (24 percent).

                          Salaries for center directors, teachers, and teacher aides make up almost
                          three-fourths of centers’ personnel costs. Other personnel costs were for
                          (1) noninstructional personnel (for example, janitors and cooks) and (2)
                          employee benefits. The average annual salary for early childhood educa-
                          tion teachers was $14,100, which was substantially less than that of
                          public school teachers and generally slightly less than that of private
                          school teachers. The average salary for teachers in urban centers was
                          $14,400 compared with a salary of $11,100 for teachers in rural centers
                          (see p. 26).


What Are the Sources of   On average, the centers received 69 percent of their income from parent
                          fees; the remainder came from federal, state, and local funds (16 per-
Centers’ Income?          cent) and other sources, such as colleges and universities, churches and
                          synagogues, and center fundraisers (15 percent).

                          The centers’ average monthly tuition fee for full-time 4-year-olds was
                          $304 per child. However, 77 (37 percent) of the 208 centers we surveyed
                          used sliding-fee schedules, which resulted in parents of low-income chil-
                          dren paying much lower monthly tuition fees (see p. 45). Of the 77 cen-
                          ters, 32 based their sliding fees on family size and income. At these 32
                          centers, for example, a family with 2 members and an annual income of
                          $8,000 paid, on average, a monthly fee of $81 for a 4-year old child. On
                          the other hand, a family of the same size with an annual income of
                          $35,000 paid, on average, a monthly fee of $280 for a 4-year old child.


                          Our analyses (see pp. 35-36) indicate that, holding other factors con-
How Is the Per-Child      stant, the annual cost per child in an early childhood education program
Cost of Early             decreases as the number of children enrolled in a center’s program
Childhood Education       increases. In addition, not surprisingly, the average cost per child
                          increases as child-staff ratios decrease.
Affected by Variations
in Center                 For example, we found that:
Characteristics?          1. A lo-percent increase in center size (measured by the number of full-
                          time equivalent children’ served) results in a less than proportional


                           ‘Includes full-time children and the full-time equivalent of part-time children.



                          Page 3                                                GAO/HRD!+O43BR        Early Childhood   Education
                       R-236126




                       increase of only 8 percent in total center operating costs if all other fac-
                       tors remain the same. Therefore, if a center with 50 children and an
                       annual per child cost of $4,500 enrolled 5 additional children (while pro-
                       viding all the additional supplies and materials needed for the new
                       enrollees), its cost per child would decrease by $82.

                       2. A reduction in a center’s child-to-staff ratio from, for example, 11 :l to
                       10: 1 (to meet the maximum allowable Smart Start criteria) results in a
                       4.6 annual percentage increase in its total operating costs. Consequently,
                       if a center with 50 children, a child-to-staff ratio of 1l:l, and a cost per
                       child of $4,500 decreased its child-staff ratio to lO:l, its annual cost per
                       child would increase by $207, to $4,707.

                       We also found that wages for teachers and teacher aides increase with
                       additional years of education or experience or both. For example, for
                       teachers, 1 year of additional education increases wages by almost 6
                       percent, and generally, 1 year of additional experience increases wages
                       by slightly more than 2 percent. In addition, centers that enroll more
                       than the average percentage of children with handicapping conditions
                       appear to pay higher wages to both teachers and teacher aides. For
                       example, a lo-percent increase in the proportion of children with handi-
                       capping conditions is associated with a 5.7-percent increase in teacher
                       wages (see p. 37).


                       In slightly more than one-third of the 208 centers surveyed, 25 percent
What Were the          or more of the children enrolled were from low-income families; about
Characteristics of     one-fourth of the centers enrolled no low-income children. Almost 70
Children Enrolled in   percent of the centers enrolled one or more children with handicapping
                       conditions. (See pp. 48-49.)
NAEYC Centers
Surveyed?




                        Page 4                                 GAO/IUtWKM3BR   Early Childhood   Education
5236126




We are sending copies of this report to other congressional committees,
the Secretary of Education, the Secretary of Health and Human Ser-
vices, and other interested parties. Please call me on (202) 275-1793 if
you or your staff have any questions. Other major contributors are
listed in appendix V.

Sincerely yours,




                                 u
Franklin Frazier
Director, Education
   and Employment Issues




Page 5                                GAO/HRD!IO-43BR   Early Childhood   Education
Contents


Letter                                                                                                      1

Section 1                                                                                                 10
Introduction                 “Smart Start” Program Provisions
                             Objectives, Scope, and Methodology
                                                                                                          10
                                                                                                          12
                             Increased Need for Preschool Care                                            14
                             Low-Income and Minority Children Less Likely to Receive                      14
                                  Preschool Services
                             Benefits of High-Quality Early Childhood Education                            15

Section 2                                                                                                  17
Costs of NAEYC               What Is the Annual Center Cost of Educating a Preschool
                                Child?
                                                                                                           17
Centers Surveyed             What Are the Centers’ Personnel Costs?                                        23
                             What Are the Centers’ Major Nonpersonnel Costs?                               32
                             What Are the Average Start-Up Costs for an Early                              34
                                Childhood Education Center?

Section 3                                                                                                  35
How Sensitive Is the         Average Cost Declines as Center Size Increases
                             Impact of Quality and Other Center Characteristics on
                                                                                                           35
                                                                                                           36
Cost of Early                     cost
Childhood Education          Indirect Influences on Total Cost: Factors That Affect                        38
                                  Wages
to Staff Salaries,           Increased Need for Teachers and Aides Will Require                            39
Quality, and Other                Slightly Higher Wages
Center
Characteristics?
Section 4                                                                                                  40
What Were the          Characteristics of Centers                                                          40
                                                                                                           46
Characteristics Of and Characteristics
                       Services        of Children
                                Provided                                                                   50
Services Provided by   Characteristics of Staff                                                            51
the NAEYC Centers
Surveyed?



                             Page 6                               GAO/HRDSO43BR   Early Childhood   Education
             Content.9




Appendixes   Appendix I: Comparison of Selected Requirements for                        60
                 Early Childhood Education Programs: S. 123 Criteria
                 and NAEYC Standards
             Appendix II: Objectives, Scope, and Methodology                            61
             Appendix III: Tables Supporting Figures in Report Text                     67
             Appendix IV: Technical Description of GAO’s Economic                       71
                 Analysis of the Costs of High-Quality Early
                 Childhood Education
             Appendix V: Major Contributors to This Briefing Report                     83

Tables       Table 1: Average Annual Cost Per Child Reported by                              2
                 NAEYC Centers (Fiscal Year 1988)
             Table 2.1: Early Childhood Education Costs at Religiously                  21
                 and Nonreligiously Sponsored Centers
                 (Fiscal Year 1988)
             Table 2.2: Average Teacher Salaries and Rent Expenses at                   21
                 Centers Serving Different Proportions of Low-Income
                 Children
             Table 2.3: Salaries of Early Childhood Education                            27
                  Teachers Compared With Those of Public Elementary
                  School Teachers (Fiscal Year 1988)
             Table 2.4: Reported Annual Costs of Supplementary                           34
                  Services
             Table 4.1: Percentage of Children With Handicapping                         49
                  Conditions at Centers
             Table 4.2: Percentage of Centers Providing                             .    50
                  Supplementary Services
             Table 4.3: Percentage of Centers Providing Specialized                      51
                  Services for Children With Handicapping Conditions
             Table III. 1: National Cost Per Child (Data for Fig. 2.1)                   67
             Table 111.2:Regional Cost Per Child (Data for Fig. 2.2)                     67
             Table 111.3:Cost Per Child by Center Sponsor (Data for                      67
                  Fig. 2.3)
             Table 111.4:Cost Per Child at Centers Serving Children of                   67
                  Different Income Levels (Data for Fig. 2.4)
             Table 111.5:Teacher Salaries in Urban Compared With                         67
                  Rural Areas (Data for Fig. 2.8)
             Table 111.6:Salaries of Early Education Teachers and                        68
                  Public and Private School Teachers by Years of
                  Experience (Data for Fig. 2.9)




              Page 7                              GAO/I3RDWBR   Early Childhood   mucation
          Contents




          Table 111.7:Salaries of Early Education Teachers and                     68
              Public and Private School Teachers by Years of
              Experience (Annualized) (Data for Fig. 2.10)
          Table 111.8:Average Directors’ and Teacher Aides’                        68
              Salaries by Region (Data for Figs. 2.7 and 2.11)
          Table 111.9:Centers Offering Certain Benefits to All or                  69
              Some Teachers (Data for Fig. 2.12)
          Table 111.10:Centers in Public and Private Sectors (Data                 69
               for Fig. 4.2)
          Table III.1 1: Center Size (Data for Fig. 4.3)                           69
          Table 111.12:Median Monthly Fees for Families by Income                  69
               (Data for Fig. 4.5)
          Table 111.13:Enrollments by Age Group (Data for Fig. 4.6)                 70
          Table III. 14: Staff Experience in Early Childhood                        70
               Education (Data for Fig. 4.14)
          Table 111.15:Reasons for Centers’ Staff Shortages (Data                   70
               for Fig. 4.15)
          Table IV. 1: Variable Definitions                                         80
          Table IV.2: Multiple Regression Cost Equation Estimates                   81
          Table IV.3 Estimates of Regression Equation for Teachers                  82
               and Aides

Figures   Figure 1,l: Four-Year-Olds in Families Below 116 Percent                  11
               of Poverty Level
          Figure 1.2: Study Objectives                                              12
          Figure 1.3: Study Methodology                                             13
          Figure 2.1: Cost Per Child Higher in Urban NAEYC                          18
               Centers
          Figure 2.2: Cost Per Child Highest in the Northeast                       19
          Figure 2.3: Cost Per Child Highest in Public NAEYC                        20
               Centers
          Figure 2.4: Cost Per Child More in Centers With No Low-                   22
               Income Children
          Figure 2.5: Salaries and Benefits Make Up Most of Center                  23
               costs
          Figure 2.6: Teaching Salaries Make Up Most of Center                      24
               Personnel Costs
          Figure 2.7: Directors’ Salaries Lowest in West for NAEYC                  25
               Centers
          Figure 2.8: Teacher Salaries Higher in Urban NAEYC                         26
               Centers




           Page8                               GAO/HWMO-MBR   Euly Childhood Education
Contents




Figure 2.9: Early Childhood Education Teachers’ Salaries:                   28
     Lower Than Those of Public School Teachers
Figure 2.10: Early Childhood Education Teachers’                            29
     Salaries: Lower Than Private and Public School
     Teachers’ (annualized)
Figure 2.11: Teacher Aide Salaries Lowest in West for                       30
     NAEY C Centers
Figure 2.12: Benefits Provided by NAEYC Centers to                          31
     Teachers
Figure 2.13: Rent and Food Make Up Almost Half of                            32
     Nonpersonnel Costs
Figure 4.1: Most NAEYC Centers Located in the South and                      41
     Midwest
Figure 4.2: Most NAEYC Centers Were Nonprofit                                42
Figure 4.3: Most NAEYC Centers Were Small or Medium-                         43
     Sized
Figure 4.4: NAEYC Centers Were Funded Mostly Through                         44
     Parent Fees
Figure 4.5: Sliding-Fee Schedules Reduced Cost for Low-                      45
     Income Children
Figure 4.6: All Centers Served 3- and 4-Year-Old Children                    47
Figure 4.7: Most Centers Served One or More Low-Income                       48
     Children
Figure 4.8: Almost Half of Directors Had Graduate                            52
     Degrees
Figure 4.9: Profile of Directors in NAEYC Centers                            53
Figure 4.10: Over Half of Teachers Had at Least                              54
     Bachelor’s Degrees
Figure 4.11: Profile of Teachers in NAEYC Centers                            55
Figure 4.12: Sixty Percent of Aides Had at Least Some                        56
     College
Figure 4.13: Profile of Teacher Aides in NAEYC Centers                       57
Figure 4.14: Staff Experience in Early Childhood                             58
      Education
Figure 4.15: Reasons for Centers’ Staff Shortages                            59



 Abbreviations

 NAEYC     National Association for the Education of Young Children
 NCES      National Center for Education Statistics
 NE4       National Education Association


 Page 9                               GAO/lSltWO4BB   Early Childhood Education
Section 1

Introduction


                         In recent years, the increase in women in the work force has resulted in
                         a growing need for child care for prekindergarten-aged children. This
                         need, coupled with the demonstrated benefits of high-quality early
                         childhood education programs, has led to numerous legislative proposals
                         for federal funding of these programs. One such measure, Smart Start:
                         The Community Collaborative for Early Childhood Development Act of
                          1989 (S. 123), would help states and local governments provide compre-
                         hensive, developmentally appropriate programs to preschool-or
                         prekindergarten-aged      chi1dren.l


                         “Smart Start” programs would be required to meet certain criteria that
“Smart Start”            are usually characteristic of high-quality programs. For example, child-
Program Provisions       teaching staff ratios could not exceed 10 to 1, and the maximum daily
                         group size (the number of children per group) could not exceed 20. Fur-
                         ther, early childhood education teachers and teacher aides would have
                         to be properly trained in early childhood education or child develop-
                         ment. Moreover, centers would be required to provide lunch to children
                         and, if requested by parents, breakfast. In addition, programs would be
                         required to provide certain supplementary services, including

                     l   health screening and screening for handicapping conditions,
                     l   information and referral for health and social services, and
                     l   parenting education, which may be conducted through such means as
                         conferences, newsletters, and orientation meetings.

                         These program criteria are also similar to the current accreditation stan-
                         dards used by the National Association for the Education of Young Chil-
                         dren (NAEYC) (see app. I). NAEYC, a membership organization of more
                         than 70,000 professionals in the field of child development and early
                         childhood education, provides the only national voluntary accreditation
                         system for early childhood education centers and schools.

                         According to S. 123, at least 67 percent of program funds made available
                         to localities with approved applications would be allocated on the basis
                         of the number of poor children under the age of 5 in each locality. Only
                         centers that provide full-day, full-year programs would be eligible to
                         receive funds. Services would be targeted primarily to 4-year-olds,
                         although 3- and 5-year-olds could also be served if all eligible 4-year-olds
                         whose parents request services were enrolled.

                         ‘A developmentally appropriate program is one that is appropriate for the child’s age and all areas of
                         the child’s development, such as educational, physical, emotional, and social.



                         Page 10                                             GAO/HRD-9043BR Early Childhood         Education
                                        Section 1
                                        Jntraduction




Figure   1 .l
r-



            GAO Four-Year-Olds in Families
                Below 115% of Poverty Level
                 Number in Thousands
                 350




                   Number of 4-Year-Olda In FamMos Below 115% of Povvorty Line (1988)

                 Note: Numbers are rounded estimates based on Census data.




                                         Under S. 123, poor children in each state would be eligible to participate
                                         without charge if their family incomes fell below 115 percent of the pov-
                                         erty leveLz According to Census Bureau data, in fiscal year 1988, more
                                         than 800,000 4-year-old children were living in such families. This popu-
                                         lation ranged from about 140,000 children in the Northeast to about
                                         350,000 in the South (see fig. 1.1). The South has both a larger poverty
                                         rate and more children than the other three regions.



                                         ‘For example, the 1988 poverty level for a family of four persons is $12,092; 115 percent of this
                                         income level is $413,906.



                                         Page 11                                             GAO/HRD-904BR        Early Chiklhood    Education
                             Section 1
                             Introduction




Figure   1.2



          G.0       Study Objectives

                l   What is the cost of providing
                    high-quality early childhood
                    education?
                l   What is the average annual
                    salary for early childhood
                    education teachers?
                .To what extent will changes in
                 certain factors (i.e., child-staff
                  ratio) affect costs?

                             To assist the Senate Committee on Labor and Human Resources in its
Objectives, Scope,and        deliberations on S. 123, the Committee Chairman asked us to determine
Methodology                  the costs of providing high-quality early childhood education. Our study
                             objectives are shown in figure 1.2.




                              Page 12                             GAO/BRD9O4BR   Early Childhood   Education
                            Section 1
                            Introduction




Figure   1.3



           GM       Study Methodology


                l   Surveyed costs and services
                    at 265 centers accredited by
                    NAEYC
                l   These centers were: full-day,
                    full-year programs serving
                    4-year-olds
                l   Response rate was 78%



                            To collect information on the costs and services at high-quality early
                            childhood education centers, we agreed to survey centers with criteria
                            similar to those specified in S. 123 (see app. I). We identified such cen-
                            ters as those full-day, full-year programs accredited by NAEYC.We
                            believe the costs of NAEYC-accredited programs would most likely be sim-
                            ilar to the costs of programs funded under the proposed Smart Start bill.
                            Our study methodology is shown in figure 1.3. Of the 265 centers we
                            surveyed, 208 (78 percent) responded to our mail questionnaire. This
                            report was reviewed by two national experts in the area of economics
                            and education, W. Norton Grubb and Henry M. Levin. (See app. II for a
                            complete description of our objectives, scope, and methodology.)



                             Page 13                              GA0/HRD9043BR   Early Childhood   Education
                         Section 1
                         Introduction




                         Traditionally, mothers have been the primary caretakers of preschool-
Increased Need for       aged children. The percentage of women with children under the age of
Preschool Care           6 who are employed has tripled from 20 to 63 percent, however, as com-
                         pared with three decades ago. This and other estimates of women’s
                         work-force participation indicate a likely increase in demand for child
                         care and educational services.

                         Care for preschool-aged children can be provided in a variety of settings,
                         from private homes to professionally operated programs, including
                         programs that have an early childhood education component. Generally,
                         early childhood education programs emphasize children’s development;
                         they are designed to support and encourage the child’s intellectual, emo-
                         tional, and social growth. In recent years, these programs have grown in
                         popularity. The percentage of children from 3 to 4 years of age enrolled
                         in early childhood education programs has risen more than threefold
                         since the mid 1960s-from 11 percent in 1965 to 39 percent in 1986.


                         Despite increased enrollments in early childhood education programs,
Low-Income and           disproportionately fewer low-income and minority children receive a
Minority Children Less   preschool education than children of higher income families. For exam-
Likely to Receive        ple, in 1984 only 33 percent of 4-year-old children in families with
                         annual incomes below $10,000 were enrolled in preschool programs,
Preschool Services       compared with 67 percent of those with annual incomes over $35,000.
                         The enrollment of white children from high-income families with non-
                         working mothers in early childhood education programs increased at a
                         greater rate than the enrollment of all other children during the period
                         1975-84.”

                         While the Department of Health and Human Services’ Head Start pro-
                         gram-the largest national public program providing educational and
                         developmental services to preschool-aged children from economically
                         disadvantaged families-served     about 450,000 children (primarily 3-
                         and 4-year-olds) in fiscal year 1988, many of the nation’s eligible chil-
                         dren remain unserved. Only one out of every six eligible low-income pre-
                         school children are served by Head Start. Many education experts believe
                         that additional funding will be needed for early childhood education
                         programs for low-income children if these children are to begin school
                         with preparation similar to that of children in higher income families.

                          “Nancy L. Kameit, “Effective Preschool Programs for Students at Risk,” in Effective Programs for
                          Students at Risk, eds. Robert E. Slavin, Nancy L. Karweit, and Nancy A. Madden (Boston: Allyn and
                          Bacon, 1989).



                          Page 14                                           GAO/BIRD-90-43BR     Early Childhood   Education
                      Section 1
                      Introduction




                      In addition, experts expect the need for early childhood education pro-
                      grams to increase in future years as the proportion of at-risk children in
                      the school system increases4 A recent study projects that the number
                      and proportion of children at risk will increase steadily from the 1980s
                      through 2020; this is because of increases in several factors that have
                      been associated with low student achievement. These include substan-
                      tial increases in the number of children (1) in poverty, (2) living with
                      only one parent, or (3) living with poorly educated mothers (those not
                      completing high school) or whose mothers’ primary language is not
                      English.”


                      Several recent research studies have demonstrated the.benefits of high-
Benefits of High-     quality early childhood education programs for children from economi-
Quality Early         cally disadvantaged families.” The Perry Preschool Program longitudinal
Childhood Education   study measured the cost-effectiveness of a high quality preschool educa-
                      tional program for 3- and 4-year-olds in Ypsilanti, Michigan. The study
                      found that about $7 is saved for each $1 invested in these programs.’
                      The study compared achievement measures in education and employ-
                      ment for disadvantaged youth (who had previously attended the Perry
                      Preschool Program) with a group of students who were similar as to
                      preprogram measures of intelligence and family socioeconomic charac-
                      teristics, but who did not attend the program. Those who attended the
                      program were found to have greater educational attainment and better
                      levels of employment. Although some benefits-such as increases in
                      scores on intelligence tests- appeared to be temporary gains, other
                       gains-such as lower retention rates in the same grade and lower place-
                      ment rates in special education classes-appeared to be longer lasting.




                       4At-risk children are those who, on the basis of several risk factors, are unlikely to graduate from
                       high school. These risk factors include low socioeconomic status, low achievement, retention in grade,
                       and poor attendance.
                       “Aaron Pallas and others, “The Changing Nature of the Disadvantaged Population: Current Dimen-
                       sions and Future Trends,” Educational Researcher, Vol. 18, No. 6 (1989), pp. 16-22.

                       “John R. Berrueta-Clement and others, Changed Lives: The Effects of the Perry Preschool Program on
                       Youths Through Age 19 (Ypsilanti, Michigan: High/Scope Educational Research Foundation, 1984).
                       Irving Lazar and Richard Darlington, Lasting Effects After Preschool: A Report of the Consortium for
                       Longitudinal Studies, DHEW Publication No. (OHDS) 7930178, (Washington, D.C.: U.S. Government
                       Printing Office, 1978). Department of Health and Human Services, Head Start Bureau, Path to -the
                       Future: Long-Term Effects of Head Start in the Philadelphia School District (Sept. 1987).
                       ‘John R. Berrueta-Clement and others, Changed Lives, p. 90.



                       Page 16                                             GAO/BRD-9043BR       Early Childhood   Education
Section 1
Introduction




The estimated cost savings from the Perry program included both the
savings from reduced costs of educational remediation classes, crime,
and unemployment and welfare payments, and the revenue generated
from taxes paid from increased earnings. One analysis indicates that
approximately 60 percent of the additional tax revenues generated as a
result of this investment would likely accrue to the federal government8

Benefits of early childhood education programs also have been docu-
mented by studies of other such programs. For example, lower grade
retention rates were found in studies of five preschool programs in Mur-
freesboro, Tennessee; New York City; New York State; Philadelphia; and
Rome, Georgia.9 Early childhood education programs also have been
found to be effective in reducing special education piacements and high
school dropout rates.




 8Henry M. Levin, “Financing the Education of At-Risk Students,” Educational Evaluation and Policy
 Analysis, Vol. 11, No. 1, (Spring 1989), p. 47.
 gSeeJohn R. Berrueta-Clement and others, Changed Lives, chapter 6. See also Department of Health
 and Human Services, Head Start Bureau, Path to the Future.



  Page 16                                          GAO/IiRD-9043BR     Early Childhood   Education
Section 2

Costsof NAEYC CentersSurveyed


                        A typical full-day, full-year NAEYC-accredited early childhood education
                        center’s annual costs are $4,200 per child, with in-kind donations
                        amounting to about another $600 per child. The centers are funded
                        chiefly through parents’ fees.

                        Personnel costs account for almost 65 percent of total costs for the cen-
                        ters surveyed. Rent and mortgage costs make up about 11 percent of
                        center costs. The remaining costs, nearly 25 percent of all costs, are for
                        educational materials and equipment, insurance, office supplies, repairs
                        and maintenance, utilities, supplementary services, food, and additional
                        operating expenses.


                        On average, early childhood education centers reported a cost per child
What Is the Annual      of $4,200 in fiscal year 1988.’ After adjusting for in-kind donations,2
Center Cost of          however, we estimate an average cost per child of about $4,800.3 Overall
Educating a Preschool   estimated costs, including the value of in-kind donations, in urban cen-
                        ters is about $1,250 more than in rural centers (see fig. 2.1). (See app. III
Child?                  for tables reporting data used in figures.)




                        ‘Although the centers we surveyed enrolled 4-yearolds, some centers also enrolled children of other
                        ages. For example, 33 percent of the centers served infants and 45 percent served toddlers. Lower
                        child-staff ratios necessary for serving infants and toddlers generally contribute to higher per child
                        costs at these centers. Although our data suggest that centers not serving infants and toddlers have
                        lower costs than centers that do, our data do not allow us to accurately determine the cost difference
                        between centers serving infants and toddlers and those not serving such children.

                        ‘On the basis of information provided by NAEYC centers, we added to their reported costs the value
                        of in-kind donations, including rent, labor, supplies, equipment, and supplementary services, such as
                        health and social services to enrolled children. For centers that reported no rent or insurance costs-
                        costs essential for operating an early childhood education center-and indicated no value for dona-
                        tions of these essential center expenditures, we estimated regional rent and insurance costs.

                        3This estimate is generally consistent with estimates from other studies on child care costs, although
                        many of these studies were limited in scope. See W. Norton Grubb, “Young Children Face the State:
                        Issues and Options for Early Childhood F’rograms,” American Journal of Education, Vol. 97, No. 4
                        (Aug. 1989), p. 379.



                        Page 17                                              GAO/BRD9043BR        Early Childhood Education
                                    Section 2
                                    Costa of NAEYC Centers   Surveyed




Figure   2.1



          GM   Cost Per Child Higher in
               Urban NAEYC Centers




                 Nathl      Urban      Runl

                 1       1 In-kind donations
                           Coat reported by NAEYC centers




                                     Page 18                            GAO/H&DMBR   Edy   Childhood   Education
                                            Section 2
                                            Costs of NAEYC Centers   Surveyed




Figure   2.2



          GAO Cost Per Child Highest
                the Northeast -
               6000   Dollars
               5500   n
               5000
               4500
               4000
               3500
               3ooo
               2500
               2000
               1500
               1000
                500
                  0




                          1     ) In-kind donations
                                  Cost reported by NAEYC centers




                                             As shown in figure 2.2, the cost of early childhood education per child is
                                             lowest in the West ($4,472) and highest in the Northeast ($5,608). In-
                                             kind donations made up between 10 percent (Midwest) and 15 percent
                                             (South) of centers’ estimated annual costs.




                                             Page 19                               GA0/HRD9043BR   Early Childhood   Education
                                                 Section 2
                                                 Costs of NAEYC Centers Surveyed




Figure   2.3



         G-0 Cost Per Child Highest in
             Public NAEYC Centers
               6000     Dollars
               5500                          ,
               5ooo
               4500
               4mo
               3!500
               3aoo




                      Ceder Spmsof

                        1         1 In-Kind Donations
                                   Costs Reportad by NAEYC Centers




Cost at Profit and at                            The estimated cost to educate a child in a public centeti is almost $1,300
                                                 more than in a for-profit center and about $650 more than in a private
Nonprofit Centers                                nonprofit center (see fig. 2.3). The greater cost per child in public cen-
                                                 ters is borne largely through in-kind donations.




                                                 “Public centers may include programs operated by federal, state, and local agencies, including public
                                                 schools. Only three centers surveyed were public-school-based. If public-school-based Centers pay
                                                 early childhood education teacher salaries that are comparable to the salaries of other public school
                                                 teachers. costs at these centers are likely to be higher than those of other programs we surveyed.



                                                  Page 20                                            GAO/HRD9O43BB        Early Childhood    Education
                                          Section 2
                                          Costs of NAEYC Centers    Surveyed




                                          In addition, annual costs per child are nearly $1,400 lower in religiously
                                          sponsored centers than in nonreligiously sponsored centers (see table
                                          2.1). The lower cost per child at religiously sponsored centers appears to
                                          be due primarily to the comparatively lower teacher salaries as well as
                                          rent and mortgage expenses. Many religiously sponsored centers receive
                                          free or subsidized rent from their church or synagogue sponsors.

Table 2.1: Early Childhood Education
Costs at Religiously and Nonreligiously                                                                          Per child
Sponsored Centers (Fiscal Year 1988)                                                                         Average
                                                                                                          annual rent      Average
                                                                                 Number of               or mortgage        teacher
                                          Center sponsorship                       centers Annual cost-’       costsa      salaries
                                          Religious         __     .-___                   37          $3,459              $398         $13,359
                                          Nonrelqious                                     168           4,824               589          14,225
                                          Tncludes In-kind donatlons and estimated rent expenses for centers that neither reported rent costs nor
                                          mdlcated a value for In-k& donatlons.




Do Costs Per Child Differ                 The estimated annual cost per child is about $500 to $900 higher at cen-
                                          ters that reported serving no low-income children than at centers that
at Centers Serving                        reported serving such children (see fig. 2.4). This cost difference
Children From
            _ - Different                 appears to be due primarily to the higher average teacher salaries and
Income Levels’!                           rent expenses at centers with no low-income children than at other cen-
                                          ters (see table 2.2).

Table 2.2: Average Teacher Salaries and
Rent Expenses at Centers Serving                                                                                                    Annual rent
Different Proportions of Low-Income                                                                                   Average      or mortgage
Children                                                                                          Number of            teacher        costs per
                                          Center                                                    centers              salary           childa
                                          Serves no low-Income children                                     46          $14,983             $491
                                          Serves 1% to 25% low-Income children                              69           14,264              371
                                                                  -.
                                          Serves more than 25% low-Income children                          63           13,396              388
                                          %cludes In-kind donabons and estimated rent or mortgage expenses for centers that neither reported
                                          rent or mortgage costs nor Indicated a value for In-kind donatlons.




                                           Page 21                                              GAO/BRIMO43BR       Early Childhood    Education
                                                    Section 2
                                                    Costs of NAEYC Centers Surveyed




Figure 2.4


        GAQ Cost Per Child More in Centers
            With No Low-Income Children
             6666       Dollars   in Thousands

             5500

             5ooo

             4500

             4ooo

             3500

             3ooo

             2500

             2ooo

             1500

             loo0

              500

                    0




                                     In-Kind   Donations




                                                       Page 22                        GAO/HRD9043BR   Early Childhood   Education
                                        Section 2
                                        Cost.6 of NAEYC Centers   Surveyed




Figure   2.5



         ~0    Salaries & Benefits Make Up
               Most of Center Costs


                                                                     Other




                                                                     Rent or Mortgage
                                                                     Salaries and Benefits




               Total canters is 150

               “Other’ indudes educational materials and equipment, insurance, office supplies, repairs and
               maintenance, utilities, health and social sarvices, food and additional operating expanses.




                                        Nearly two-thirds of center costs go to salaries and benefits (see fig.
What Are the Centers’
Personnel Costs?




                                         Page 23                                             GAO/~90-43BB     Early Childhood   Education
                                       Section 2
                                       Costs of NAEYC Centers   Surveyed




Figure   2.6



           GJB Teaching Salaries Make Up
               Most Center Personnel Costs

                                                                   Employee     Benefits


                                                                   Other   Salaries




                                                                   Teaching    Salaries




                Total centers is 150




                                        On average, salaries for instructional staff make up 74 percent of cen-
                                        ters’ personnel costs. Salaries for noninstructional personnel make up
                                         13 percent of personnel costs. An additional 13 percent is spent on
                                        employee benefits, including employer contributions for Social Security
                                        and Medicare (see fig. 2.6).




                                        Page 24                                            GAO/‘IiRDQO-43BR   Jkrly   Childhod   Education
                      Section 2
                      Costs of NAEYC Centers   Surveyed




Figure   2.7



          GAO Directors’ Salaries Lowest
              in West for NAEYC Centers




Director Salaries      The average annual salary for directors of early childhood education
                       centers in our survey was $24,340. On average, directors’ salaries were
                       lowest in the West at $21,500. In each of the other three regions, the
                       average annual salary for center directors was about $25,000 (see
                       fig. 2.7).




                       Page 26                              GAO/lIRIMO4BB   Early Childhood   Education
                                        Section 2
                                        Costs of NAEYC Centers Surveyed




Figure   2.6



         GAO Teacher Salaries Higher in
             Urban NAEYC Centers
               l!SOO
               14000
               13000
               12000
               llooo
               10000
                9ooo
                8ooo
                7ooa
                6ooa
                5ow
                4ooa
                3ooo
                2ooo
                1000



                       Nation   Urban      RUd




Teacher Salaries                         We estimate that the fiscal year 1988 average annual salary for early
                                         childhood education teachers at NAEYC centers was about $14,100.5
                                         Teacher salaries, on average, were substantially higher in urban centers
                                         than in rural centers. The average salary for teachers in urban centers
                                         was $14,400; the average salary in rural centers was $11,100 (see
                                         fig. 2.8).



                                         “This estimate is slightly higher than that reported for child care teaching staff with Cyear college
                                         degrees in the Child Care Employee Project’s 1989 National Child Care Staffing Study.



                                         Page 26                                              GAO/IiRD-9043BR      Early Childhood    Education
                                         Section 2
                                         Costa of NAEYC Centers     Surveyed




                                         The average annual salary for teachers varied across regions, ranging
                                         from $12,900 in the West to $15,500 in the Northeast. As shown in table
                                         2.3, in each region, salaries of early childhood education teachers are
                                         approximately half that of public elementary school teachers.

Table 2.3: Salaries of Early Childhood
Education Teachers Compared With                                                                                           Teachers
Those of Public Elementary School                                                                                         Early       Public
Teachers (Fiscal Year 1988)                                                                                         childhood    elementary
                                         Region                                                                     education        school’
                                         Northeast                                                                     $15,500          $30,200
                                         Midwest                                                                        14,100           27.800
                                         South                                                                          14,200           24,500
                                         West                                                                           12,900           29,600
                                         3ource    National Education Assoclatlon, Estimates of School Statistics: 1987-88


                                         Using salary data from a recent survey report from the National Center
                                         for Education Statistics (NCFS), we also compared the salaries of early
                                         childhood education teachers with those of public and private school
                                         teachers” with varying experience (see fig. 2.9). When comparisons were
                                         made across groups with similar years of experience, the differences
                                         between the salaries of early childhood education teachers and public
                                         school teachers narrowed somewhat. The public and private school
                                         teacher salaries included in figure 2.9 are based on a school year of
                                         approximately 10 months.




                                         “These include both elementary and secondary school teachers because separate data were not avail-
                                         able for private elementary school teachers.



                                         Page 27                                               GA0/HRD9043BR         Early Childhood   Education
                                               Section 2
                                               Costs of NAEYC Centers                Surveyed




Fiaure 2.9


       MO    Early Childhood Education
             Teachers’ Salaries:
             Lower than those of public
             school teachers
             32
                                                                               --000~~00----    -0”---,
             26                                                     ---00---
                                                        0-L   0-c
                                                 -I--
             24                          -4-
                       ___m----   #---
             20 me--

             16

             12

              6

              4




                                               NCEShas noted that private school salaries reported do not include in-
                                               kind income. It found that 23 percent of private school teachers earned
                                               an average in-kind income of $2,900, which may include housing, meals,
                                               transportation, and reduced tuition rates. In-kind income does not apply
                                               to public school teachers. In addition, NCES found that in private schools,
                                               lay teachers (those who are not from a religious order) earn salaries that
                                               are higher than those paid to teachers who are members of a religious
                                               order. According to NCES, approximately 1 in 10 private school teachers
                                               surveyed is a member of a religious order.




                                                Page 28                                                   GAO/HRD-9043BR   Early Childhood   Education
                                                          Section 2
                                                          Costs of NAEYC Centers Surveyed




Figure 2.10


       GAQ Early Childhood Education
           Teachers’ Salaries:
              Lower than private & public
              school teachers’ (annualized)
                                                                                               **---   ****.****------
              35
                                                                                      ***---
              32                                                               **--
                                                                         -0*
                                                             .c--
              25                                   *cc-
                                          ***--
                           ***---
              24 -0--

              20

              10

              12

               5

               4

               0
                                                                                                                         \




                    -               Early Ctddbal    Edmkm          T-
                    ----            PticScboiTeccher4
                    -               Pnvate school Teachen




                                                             Across all levels of experience, public school teacher salaries are sub-
                                                             stantially higher than those of early childhood education teachers. Early
                                                             childhood education teacher salaries were roughly similar, however, to
                                                             those of private school teachers. Figure 2.10 compares annualized sala-
                                                             ries (assuming that teachers are working 12 months rather than 10) for
                                                             public and private school teachers. Most early childhood teachers
                                                             worked 12 months, and salaries have been annualized for the others.




                                                             Page 29                                                         GAO/HRD-9043BR   Early Childhood   Education
                                                  Section 2
                                                  Costa of NAEYC Centers Surveyed




Figure 2.11


       MO     Teacher Aide Salaries Lowest
              in West for NAEYC Centers
              15000         Dollam
              14000
              13OW
              12000
              lloca
              loo00                       I
               woo                                                \
               8ooa
               7ooo
               6wo
               5ooo
               4ooo
               3wo
               2ooo
               1000
                      o\


                       9$             J           $8         g        J

                                 9’           *




Teacher Aide Salaries                              Teacher aides in the centers we surveyed earned, on average, salaries of
                                                   approximately $10,200. Regional variations in aides’ salaries were simi-
                                                   lar to those for directors, with teacher aide salaries lowest in the West
                                                   and about the same in the other three regions. As shown in figure 2.11,
                                                   teacher aide salaries in the four regions were, on average, as follows:
                                                   $9,200 in the West, $10,200 in the Northeast, $10,400 in the Midwest,
                                                   and $10,600 in the South.




                                                   Page 30                              GAO/HRD-!30-43BR   Early Childhood   Education
                              Section 2
                              Costs of NAEYC Centers   Surveyed




Figure 2.12


        GAO Benefits Provided by NAEYC
            Centers to Teachers




                 I   SomeTeachers
                     All Teachers




Staff Benefits                 On average, centers spent approximately $416 per child on staff bene-
                               fits, which make up about 15 percent of total staff salaries. Estimated
                               staff benefit costs include employer contributions for Social Security
                               and Medicare coverages. The extent to which other benefits are offered
                               to teachers is shown in figure 2.12.

                               A vast majority (about 88 percent) of centers offered paid sick leave
                               and vacation leave to all their teachers. Health insurance was offered to
                               all teachers in 65 percent of all centers. Less than 50 percent of centers
                               offered such benefits as pension coverage, life insurance, and reduced
                               child care fees to its teachers.


                                Page 31                              GAO/BRDW43BR   Early Childhood   Education
                                         Section 2
                                         Costs of NAEYC Centers      Surveyed




Figure   2.13



          GM    ‘Rent and Food Make Up Almost
                Half of Nonpersonnel Cbsts
                                                                        Other




                                                                        Rent or Mortgage




                                                                        0%
                                                                        Educational    Materials    and Equipment




                Totelcentersis   113

                Other includes telephone and utilities, repairs end maintenance,   office supplies and equipment,
                center insurance, health and social services, end miscellaneous    expenses.




                                         Of centers’ nonpersonnel costs, rent or mortgage costs made up slightly
What Are the Centers’                    more than one-fourth; food costs made up almost one-fifth; educational
Major Nonpersonnel                       materials and equipment (which may also include the cost of field trips)
Costs?                                   made up approximately 8 percent; and the costs of insurance, office sup-
                                         plies, repairs and maintenance, utilities, supplementary services, and
                                         additional operating expenses made up the other 45 percent (see fig.
                                         2.13).




                                          Page 32                                                  GAO/HRD9O43BR    Early Childhood   Education
                         Section 2
                         Costs of NAEYC Centers   Surveyed




Rent or Mortgage Costs   Out-of-pocket rent or mortgage costs averaged $338 per child. When the
                         value of donated space is also considered, the average rent or mortgage
                         costs increased to $532 per child. After estimating rent or mortgage
                         costs for centers that did not report either out-of-pocket expenses or
                         donations of space, the average rent or mortgage cost was $555 per
                         child.


Food Costs               In fiscal year 1988, centers spent an average of $255 per child for food.
                         Costs were higher-at $300 per child-for     centers that serve, at a mini-
                         mum, breakfast and lunch.


Insurance Costs          Overall, centers paid an average of $90 per child for insurance in fiscal
                         year 1988. Insurance costs include primarily coverage for liability (such
                         as bodily injury, property damage, and personal injury), real property,
                         personal property, theft, and employee dishonesty. The average per-
                         child cost of insurance decreased as the size of the center increased. For
                         example, annual insurance costs for centers with 50 or fewer children
                         averaged $102 per child compared with $84 per child for centers with
                         more than 100 children.




                         Page 33                               GAO/BRiMO43BR   Early Childhood   Education
                                      Section 2
                                      Costs of NAEYC Centers Surveyed




Costs of Supplementary                On average, each center spent $151 per child in fiscal year 1988 for sup-
                                      plementary services7 See table 2.4 for the average costs of each of these
Services                              services.

Table 2.4: Reported Annual Costs of
Supplementary Services                                                                                                           Number of
                                                                                                               Average              centers
                                                                                                               cost per           reporting
                                      Supplementary services                                                      child               costs
                                      Parent education and family support                                              $7           ___ 106
                                      Information and referral for health and social services                          22                34
                                      Health screeninga                                                                15                67
                                      Screening for handicapprng -- conditions                                         14                22
                                      Mental health services                                                           21                27
                                      Social services                                                                  16                17
                                      Transportation                                                                   56                47
                                      Total                                                                         $151
                                      9ncludes speech, language, hearing. and wsion tests; medical treatment; and preventive care



                                      Of the 27 accredited centers that had been in operation less than 5
What Are the Average                  years, 24 (89 percent) provided information on start-up costs. These
Start-Up Costs for an                 costs ranged from $8,000 to $900,000, with a median of $48,500, and
Early Childhood                       included, in descending order of amount spent, costs for space, supplies
                                      and equipment, planning and administration, and teacher training.
Education Center?




                                      7The costs for supplementary services may be underestimated. Ekcause many centers do not sepa-
                                      rately budget or account for expenses related to supplementary services, such costs may be included
                                      under a center’s educational expenses account or staff salaries account. In addition, NAEYC center
                                      costs for supplementary services may not be reflective of the costs of supplementary services under
                                      S. 123, because centers may not provide services to the extent intended by S. 123.



                                       Page 34                                            GAO/HRD-!XM3BR       Early Childhood     Education
Section 3

How SensitiveIs the Cost of Early Childhood
Education to Staff Salaries,Quality, and
Other Center Characteristics?
                        On the basis of our analysis, we conclude that (1) the cost per child
                        decreases significantly as the number of children enrolled in a center
                        increases and (2) tightening of quality standards for early childhood
                        education centers will likely increase the cost per child, either directly,
                        as a result of centers’ hiring additional teachers to lower the child-staff
                        ratio, or indirectly, as a result of centers’ paying higher wages to attract
                        better qualified teachers and aides. However, the “quality” standards
                        specified in S. 123 should increase the cost per child only of early child-
                        hood education centers that do not currently meet such standards. In
                        addition, an increase in the number of children attending centers that
                        would result from the enactment of S. 123 will necessitate hiring addi-
                        tional teachers. Evidence from previous economic research suggests that
                        this need for additional teachers will cause wages (and hence average
                        cost) to rise moderately. (See app. IV for a description of our economic
                        cost model and detailed analysis of results.)


                        Our regression results indicate that as the number of full-time equiva-
Average Cost Declines   lent children enrolled’ in a center increases, the cost per child decreases.
as Center Size          That is, a lo-percent increase in center size results in an increase of only
Increases               8 percent in total cost if all other factors remain the same.? For example,
                        if a center with 50 children and a cost per child of $4,500 enrolled 5
                        additional children (while maintaining the original child-staff ratio and
                        providing the necessary additional supplies and space), its cost per child
                        would decrease by $82. 1

                        These results must be interpreted with care, as the implied cost advan-
                        tage of large centers may be overstated for two reasons. First, if the
                        administrative burden on each director increases with center size, larger
                        centers might have to offer higher salaries than those paid at smaller
                        centers to attract capable administrators. Higher salaries would at least
                        partially offset the cost per child differential between small and large
                        centers.

                        Second, an increase in center size achieved by the consolidation of many
                        small centers into fewer larger centers would impose additional costs on
                        parents. For example, the decrease in the number of centers could

                        ‘Includes full-time children and the full-time equivalent of part-time children

                        ‘The standard errors of the estimates discussed in this section are reported in table IV2

                        ‘Cost per child declined with increases in center size even when we restricted our analysis to centers
                        with more than 68 children. See footnotes 8 and 9 in app. IV for a discussion of the economies-of-scale
                        issue.



                        Page 35                                              GAO/HRS9043BR         Early Childhood Education
                        Section 3
                        How Sensitive Is the Cost of Early Childhood
                        Education to Staff Salaries, Quality, and
                        Other Center Characteristics?




                        increase the average commuting distance for parents and children.
                        Accordingly, center size may be limited by local conditions-such  as
                        population density-that    affect the number of children who can be
                        served by a given center:&

                        Because quality standards in early childhood education are not directly
                        measured in our study, we cannot determine from our data whether
                        quality is higher or lower in larger centers. Our analysis, however,
                        included several variables generally thought to be associated with qual-
                        ity-average    group size, child-to-staff ratio, and the wages of the direc-
                        tors, teachers, and teacher aides.” Nonetheless, when these variables are
                        held constant, average cost decreases as center size increases.”



Impact of Quality
and Other Center
Characteristics
on Cost

Child-to-Staff Ratios   Our regression results show that a reduced child-to-staff ratio increased
                        total costs for the centers in our survey. A decrease of one child per
                        teaching staffmember increased costs by 4.6 percent. That is, if a center
                        with 50 children, a child-staff ratio of 11 to 1, and a cost per child of
                        $4,500 reduced its child-staff ratio to 10 to l-without    increasing the
                        number of children enrolled in the center-its cost per child would
                        increase by about $207, resulting in a cost per child of $4,707.’



                        ‘Low population density does not necessarily mean that, on average, rural centers will be smaller
                        than urban centers. In our survey of 208 centers, the average number of full-time equivalent children
                        was nearly identical for rural and urban centers (83 and 81, respectively). See footnote 11 in app. IV.

                        ‘For example, in a 1978 study of day care costs, Abt Associates found group size to be negatively
                        correlated with quality. In addition, wages are positively correlated with the education and experi-
                        ence of directors, teachers, and aides.

                        “If the larger centers have been in existence longer than smaller centers. it is possible that some of the
                        cost reduction may be due to efficiency gained through experience in operation, rather than econo-
                        mies of scale. Our results indicate, however, that this is unlikely; total costs for centers that have
                        been in operation for less than 5 years are not statistically different from centers that have been in
                        operation longer.

                        ‘The additional staffing requirement could be met by hiring part-time teachers or teacher aides



                        Page 36                                               GAO/HRD-9043BR        Early Childhood    Education
                               Section 3
                               How Sensitive Is the Cost of Early Childhood
                               Education to Staff Salaries, Quality, and
                               Other Center Characteristics?




                               Conversely, an increased child-to-staff ratio resulted in lower total costs
                               for the centers in our survey.

                               Although one might expect that increased child-to-staff ratios would
                               result in higher wages (because of the greater responsibility required of
                               each staff member)-and      hence greater center costs-we found no evi-
                               dence of this from our analyses. Instead, the only effect of increased
                               ratios seems to be a center’s saving in the amount of teaching staff
                               needed (with its concomitant salary saving).


Children’s Group Size          There is no statistical evidence to suggest that group size (the number of
                               children assigned to a particular classroom or group in the center)
                               affects average cost.


Costs for Children With        Centers that enrolled higher percentages of children with handicapping
Handicapping Conditions        conditions tend to pay higher wages to both teachers and aides. A lo-
                               percentage-point increase in the number of these children increases
                               wages by 5.7 percent for teachers and 1.9 percent for aides.


For-Profit Centers             We estimate that for-profit centers pay wages that are 3 percent lower
                               for teachers and 7.2 percent lower for aides (even after controlling for
                               the education and experience levels of those staff) than nonprofit cen-
                               ters. No evidence suggests, however, that for-profit centers are any
                               more efficient than nonprofit centers. That is, after controlling for sala-
                               ries, occupancy cost, and the cost of supplies, the total costs of for-profit
                               centers are roughly equivalent to those of nonprofit ones.


Cost Differentials   Because   The cost of early childhood education is lower in the West, Midwest, and
of Location                    South than in the Northeast, in part because salaries for teachers and
                               aides are lower in those regions: in the West, roughly 14 percent lower
                               for teachers and 8 percent lower for aides; in the Midwest, about 19 per-
                               cent lower for teachers and 9 percent lower for aides; and in the South,
                               about 17 percent lower for teachers and 5 percent lower for aides, even
                               when other factors that our analysis showed affect salaries (such as
                               experience and education) are held constant.

                               Salaries are about 20 percent higher for teachers and 9 percent higher
                               for aides in urban areas than in nonurban areas. After controlling for



                               Page 37                                        GAO/HRLMO43BR   Early Childhood   Education
                           Section 3
                           How Sensitive Is the Cost of Early Childhood
                           Education to Staff Salaries, Quality, and
                           Other Center Characteristics?




                           other costs, the total costs for centers in urban areas are about 7 percent
                           more than for centers in nonurban areas.

                           Wages for both teachers and aides appear to be lower at centers with
                           higher proportions of children from low-income families. For example,
                           on average, a lo-percentage-point increase in the proportion of low-
                           income children is associated with wages that are approximately 2 per-
                           cent lower for both teachers and aides. This may reflect demand condi-
                           tions (less demand from low-income families for early childhood
                           education and teachers and/or less ability to pay), supply conditions
                           (less costly to hire workers in depressed areas), or both. After adjusting
                           for the impact on wages, the proportion of low-income children does not,
                           however, have a significant effect on total cost of operating a center.



Indirect Influences on
Total Cost: Factors
That Affect Wages

Wages Rise With            Wages for staff-both     teachers and teacher aides-rise with additional
Education and Experience   years of education or experience or both. On average, each additional
                           year of education increases wages by 6.0 percent for teachers and 3.3
                           percent for aides. Additional years of experience also increase wages,
                           although the percentage increase declined with years of experience. The
                           average teacher reported 5.8 years of experience; an additional year of
                           experience would increase wages by 2.3 percent. The average aide has
                           3.1 years of experience; an additional year of experience would increase
                           wages by 1.9 percent.


Center Size and Wages      Larger centers-as measured by the total number of directors, teachers,
                           and aides-tended to pay higher wages to teachers, but lower wages to
                           aides, even after adjusting for the staff’s education and experience.


A Higher Minimum Wage      The increase in the federal minimum wage from $3.35 to $3.80, slated to
                           occur in April 1990, will raise the average wage for teachers and teacher
                           aides. In fiscal year 1988, at the centers we surveyed, 4.5 percent of the
                           teachers and 13.9 percent of the aides reported earning less than $3.80
                           per hour. However, the percentage of child care teachers and teacher



                           Page 38                                        GAO/HRD9043BR   Early Childhood   Education
                         Section 3
                         How Sensitive Is the Cost of Early Childhood
                         Education to Staff Salaries, Quality, and
                         Other Center Characteristics?




                         aides affected by the increase in the federal minimum wage could be
                         even larger if centers attempt to maintain relative pay scales by also
                         raising the wages of teaching staff making more than $3.80 per hour.


Link Between Wages and   A lo-percent increase in teachers’ hourly wages causes a center’s total
                         cost to increase by 3 percent. The same increase in aides’ hourly wages
Total Center Cost        causes a 1.3-percent increase in total center costs. An increase of 10 per-
                         cent in directors’ monthly wages caused an increase of 1.7 percent in
                         total center costs.


                         An expansion in the number of children attending early childhood edu-
Increased Need for       cation centers will require increased wages to attract additional teachers
Teachers and Aides       and aides to the field. Although we made no attempt to estimate the
Will Require Slightly    cost, previous economic research on teachers suggests that moderate
                         wage increases attract many new workers to the field. If quality stan-
Higher Wages             dards for teachers, such as academic achievement, remain unchanged, a
                         lo-percent increase in wages is likely to result in a 24- to 32-percent
                         increase in the number of teachers.” Because aides have less formal
                         training than teachers, the same percentage increase in aides’ wages
                         would most likely result in an even larger percentage increase in the
                         number of aides. However, the actual wage increase caused by the
                         enactment of S. 123 or other child care bills will depend on the number
                         of new teachers required, the availability of qualified personnel, and
                         other institutional factors-such   as the degree to which early childhood
                         education workers are covered by collective bargaining agreements.




                         ‘Charles F. Manski, “i\cademic Ability, Earnings, and the Decision to Become a Teacher: Evidence
                         from the National Longitudinal Study of the High School Class of 1972,” in Public Sector Payrolls,
                         ed. David A. Wise (Chicago: University of Chicago Press, 1987).



                          Page 39                                            GAO/HRD4043BR        Early Childhood   Education
What Were the Characteristicsof and Services
Provided by the NAEYC CentersSurveyed?

                         According to our survey of KAEYC-accredited centers, a typical early
                         childhood education center is urban, nonprofit, and located in the South
                         or Midwest, and serves about 80 children.

                         The KAEYC centers surveyed generally had the same characteristics as
                         the proposed Smart Start centers, discussed in S. 123. In some cases
                         (child-staff ratio, daily group size, meals served, and teachers’ and
                         teacher aides’ training), the NAEYC centers exceeded the bill’s criteria for
                         high-quality early childhood education. All NAEYCcenters surveyed had
                         full-day, full-year programs serving 4-year-olds. The typical center also

                     l had a g-to-1 child-staff ratio for 4-year-olds;
                     . had a daily group size of 17 for 4-year-olds;
                     l provided lunch and snacks and, in some cases, lunch and breakfast;
                     l employed directors with bachelor’s or graduate degrees and 15 years of
                       experience;
                     l employed teachers, most of whom had bachelor’s degrees and 6 years of
                       experience;
                     l employed teacher aides with at least some college training and 3 years
                       of experience; and
                     l provided parent education and information or referrals for health and
                       social services.

                         In addition, most of the centers enrolled one or more low-income chil-
                         dren, although only 12 percent of the centers enrolled 50 percent or
                         more low-income children. The total average monthly fee for 4-year-olds
                         attending on a full-time basis was $304, with 37 percent of the centers
                         using sliding-fee schedules, which permitted low-income families to pay
                         much lower fees for their children.



Characteristics of
Centers

Where Were Centers       Most of the centers that responded to our survey (88 percent) were in
                         urban areas. As shown in figure 4.1, most of the NAEYC centers surveyed
Located?                 were in the South or Midwest.




                         Page 40                                GAO/HRD-9043BR   Early Childhood   Education
                                      Section 4
                                      What Were the Characteristics of and
                                      Services Provided by the NAEYC
                                      Centers Surveyed?




Fiaure   4.1



          GAO Most NAEYC Centers Located
              in the South and Midwest
                                                                 South



                                                                 Northeast




               r       33%



                                                                 Midwest



                          18%



               ‘c:
               Total Centers is 208
                                                                 West




                                      Page 41                                GAO/HRD9043BR   Early Childhood   Education
                              Section 4
                              What Were the Characteristics  of and
                              Services F’covided by the NAEYC
                              Centers &uveyed?




Figure   4.2



          GAO Most NAEYC Centers
              Were Nonprofit
               80   Percent




To What Extent Were           Eighty-five percent of the centers surveyed were private nonprofit or
Centers Publicly or           public; the others were private for-profit (see fig. 4.2). In addition, of
                              the 208 centers that responded to our questionnaire, 18 percent were
Privately Operated?           sponsored by churches, synagogues, or other religious organizations.




                               Page 42                                GAO/HRD9@43BR   Early Childhood   Education
                                                                                                                                                 -
                                            Section 4
                                            What Were the Characteristics of and
                                            Services Provided by the NAEYC
                                            Centers Surveyed?




Figure   4.3



         GAO Most NAEYC Centers Were
             Small or Medium-Sized
               80   Percent


               50


               40


               30


               20


               10


                0

                              Medium       Large    VW
                                                    l-we




How Many Children Were                      On average, in the centers we surveyed, 82 children were enrolled.1
Enrolled in Each Center?                    NAEYCclassifies center sizes according to the following guidelines:

                                       l Small: Center with fewer than 60 children enrolled.
                                       l Medium: Center with 60 to 120 children enrolled.
                                       . Large: Center with 121 to 240 children enrolled.
                                       . Very Large: Center with more than 240 children enrolled.
                                             Based on these guidelines, as shown in figure 4.3, most of the centers we
                                             surveyed were small and medium-sized.

                                             ‘Includes full-time children and the full-time equivalent of part-time children



                                             Page 43                                              GAO/HRD-9043BR        Early Childhood   Education
                                        Section 4
                                        What Were the Characteristics of and
                                        Services Provided by the NAEYC
                                        Centers Surveyed?




Figure   4.4
r
          GAO NAEYC Centers Were Funded
              Mostly Through Parent Fees
                    11                                                Local,   State,   and Federal   Funds

                    I      I                                          Other




                                                                      Parent    Fees




               Total Centers is 202

               Note: Other sources of income indude institutions of higher education,    churches or synagogues,
               employers, grants and community donations, and center fundreising.


L




What Were the Centers’                   NAEYCcenters responding to our survey received 69 percent of their rev-
Revenue Sources?                         enues from parent fees; other center revenues came from federal, state,
                                         and local funds, as well as other sources, such as colleges and universi-
                                         ties, churches and synagogues, and center fundraising (see fig. 4.4).




                                         Page 44                                                GAO/HRD9043BR      Early Childhood   Education
                                           Section 4
                                           What Were the Characteristics of and
                                           Services Provided by the NAEYC
                                           Centers Surveyed?




Fiaure   4.5



          GAO Sliding-Fee Schedules Reduced
              Cost for Low-Income Children
                300   MedianCost




                  6ooo                        16000                  25ooo                   4woo                    75000
                  Family’s Annual Income

                         -      Family of 2
                         -1-1   Family of 4
                         m      Family of 6




Monthly Fees/                                 Overall, the average monthly fee for a full-time 4-year-old was $304. At
                                              the centers that adjust fees according to annual family income and fam-
Fee Schedules                                 ily size,” however, the monthly fee for low-income children was much
                                              less. For example, the median monthly fee for the 32 centers that
                                              adjusted fees according to family size and income was $34 for children
                                              from families that (1) have four members and (2) earn no more than
                                              $8,000 annually (see fig. 4.5).



                                              “Of the 208 centers surveyed, 77 (37 percent) reported using sliding-fee schedules. Of these 77 ten-
                                              ters, 32 adjusted fees according to both family income and size.



                                              Page 45                                               GAO/HRDSO43BR       Early Childhood   Education
                        Section 4
                        What Were the Characteristics of and
                        Services Provided by the NAEYC
                        Center8 Surveyed?




What Meals Do Centers   All of the 208 centers surveyed provided at least one meal or snack. Of
Provide?                the centers surveyed, 80 percent provided at least lunch; 58 percent pro-
                        vided at least breakfast and lunch. Virtually all of the centers provided
                        an afternoon or morning snack.



Characteristics of
Children

What Age Groups Are     The centers surveyed enrolled a total of 21,417 children, ranging in age
Served?                 from infants (aged O-l 2 months) to children aged 5 and older. Children
                        aged 3 and 4 were enrolled by virtually all of the centers we surveyed
                        (see fig. 4.6). Only 33 percent of the centers enrolled infants, and 45
                        percent of the centers enrolled toddlers (aged 1 to less than 2 years).




                        Page 46                                GAO/HRDS@43BR   Early Childhood   Education
                                               Section 4
                                               What Were the Characteristics of and
                                               Services Provided by the NAEYC
                                               Centers Surveyed?




                                                                                                     -
Fiaure   4.6



           GM   All Centers Served 3- and
                4-Year-Old Children
                loo        Potwnt of Centers

                 90

                 80

                 70

                 60

                 50

                 40

                 30

                 20

                 10

                      0




                          Agm Groups ot Enrolled Children




                                                 Page 47                              GAO/HRMO43BR       Early Childhood   Education
                                       Section 4
                                       What Were the Characteristics of and
                                       Services Provided by the NAEYC
                                       Centers Surveyed?




Figure   4.7
I


           GAO Most Centers Served One or
               More Low-income Children
                                                                  26% or more




                                                                   None




                Total centers is 178




L                                                                                                                             J
What Proportion of                     Slightly more than one-third of the centers counted 25 percent or more
Children Served Are Frmom              of their enrollees as low-income children (see fig. 4.7). About one-fourth
                                       of the centers enrolled no low-income children. In addition, 44 centers
Low-Income Families?                   (25 percent of all centers) reported serving 50 percent or more low-
                                       income children. Overall, 21 percent of the children enrolled in the cen-
                                       ters surveyed were identified by center directors as being from low-
                                       income families.




                                       Page 48                                  GAO/HRD9043BR   Early Childhood   Education
                                         Section 4
                                         What Were the Characteristics of and
                                         Services Provided by the NAEYC
                                         Centers Surveyed?




How Many Children With                   Almost 70 percent of the 208 centers enrolled a total of about 900 chil-
                                         dren with handicapping conditions; the other 30 percent enrolled no
Handicapping Conditions                  such children. The number of children with handicapping conditions
Are Enrolled?                            attending a center ranged from 1 to 78. As shown in table 4.1, centers
                                         served children with a variety of handicapping conditions, including
                                         developmentally delayed, speech-impaired, and emotionally disturbed
                                         children.

Table 4.1: Percentage of Children With
Handicapping Conditions at Centers       Flnures In oercents
                                                                                                                    Centers
                                                                                                                     serving
                                         Handicapping condition                                                     children
                                         Developmentally delayed                                                           54
                                         Speech Impaired                                                                   50
                                         Emotlonallv dlsturbed                                                             31
                                         Orthopedically ImpaIred                                                           17
                                         Visually Impaired                                                                 16
                                         Hearing impaired or deaf                                                          16
                                         Mental retardation                                                                15
                                         Multiple handicaw                                                                 11
                                         Deafness and blindness                                                             1
                                         Other health ImpaIred                                                              1




What Is the Average                      On average, the centers we surveyed reported a g-to-1 child-staff ratio
                                         for 4-year-old children. The average child-staff ratio for all age groups
Child-Staff Ratio and                    was8 to 1.
Group Size?
                                         The average daily group size was 17 for 4-year-old children and 14 for
                                         all age groups.




                                          Page 49                               GAO/HRD-90-43BR   Early Childhood   Education
                                   Section 4
                                   What Were the Characteristics of and
                                   Services Provided by the NAEYC
                                   Centers Surveyed?




Services Provided

What Supplementary                 More than 85 percent of the 208 centers reported providing (1) parent
                                   education in the form of conferences, newsletters, and orientation meet-
Services Are Provided at           ings and (2) information on and referrals for health and social services
NAEYC Centers?                     for the child and family-two      of the four supplementary services speci-
                                   fied in S. 123 (see table 4.2). Less than 25 percent of the centers
                                   reported providing screening for handicapping conditions and health
                                   screening-the other two services specified in S. 123.

                                   In addition, slightly more than one-fifth of the centers provided mental
                                   health and social services to children attending their center-services
                                   not specified in S. 123 (see table 4.2). Of the centers surveyed, 15 per-
                                   cent provided transportation services, which also are not specified in
                                   S. 123.

Table 4.2: Percentage of Centers
Providing Supplementary Services   Figures in percents
                                                                                                                                 Centers
                                                                                                                                providing
                                   Services that S. 123 would require:                                                            service
                                   Parent education and famrly support                                                                  87
                                   InformatIon and referrals for health and social services                                             86
                                   Screening for handrcapplng condrtions                                                                24
                                   Health screenInga                                                                                     19

                                   Services that S. 123 would not require:
                                   Mental health services                                                                                22
                                   Social services                                                                                       21
                                   Transportabon                                                                                         15
                                   Tn ad&ion, (1) man tests and (2) speech, language, and hearrng tests were provided by 58 percent
                                   and 67 percent of the centers, respectively Although not specified In S 123, these tests are considered
                                   types of health screenmg services




                                   Page 50                                              GAO/HRD-90-43BR       Early Childhood   Education
                                        Section 4
                                        What Were the Characteristics of and
                                        Services Provided by the NAEYC
                                        Centers Surveyed?




What Services Are                       Seventy-one percent of the centers that enrolled children with handicap-
                                        ping conditions provide specialized services for these children (see table
Provided for Children                   4.3). Of these centers, 81 percent reported that, at a minimum, they pro-
With Handicapping                       vide supplementary services, such as transportation, speech therapy,
Conditions?                             physical therapy, and counseling. Other services provided by the centers
                                        included special classroom materials and equipment, such as wheelchair
                                        ramps, specialized teacher training, and teacher aides.

Table 4.3: Percentage of Centers
Providing Specialized Services for      Figures in percents
Children With Handicapping Conditions                                                                                                Centers
                                                                                                                                    providing
                                        Service provided                                                                              service
                                        Supplementary services                                                                              81
                                        Specialized teacher training                                                                        48
                                        Special classroom and building materials
                                                                             -                                                      -.__    20
                                        Teacher aides                                                                                       20




Characteristics
of Staff

Staff Education and                     Overall, educational staff at the NAEYCcenters, including directors,
Experience Levels                       teachers, and teacher aides,:3had specialized training or experience in
                                        early childhood education.




                                        “For the purposes of our questionnaire, we defined educational staff members as the following: direc-
                                        tor-a person who has primary responsibility for administering the program. which may also include
                                        teaching responsibilities; teacher-a person in charge of a group of children, often with staff supelvi-
                                        sory responsibilities; teacher aide-a person working under the supervision of a teacher who helps
                                        with the care and education of a group of children.



                                        Page 51                                             GAO/HRD4043BR         Early Childhood   Education
                                        Section 4
                                        What Were the Characteristics of and
                                        Services Provided by the NAEYC
                                        Centers Surveyed?




Figure   4.8



          GAO Almost Half of Directors
               Had Graduate Degrees
                              1                                   Eociate’s    Degrees
                                                                  9%
                                                                  Other




                                                                  Graduate Degrees




                          c                                       Bachelor’s Degrees
               Total directors is 207




Directors                               Almost half of all early childhood education center directors reported
                                        having graduate degrees (see fig. 4.8) with most in early childhood edu-
                                        cation. Another 41 percent reported having bachelor’s degrees; the other
                                        directors reported having associate’s degrees or other training. Figure
                                        4.9 illustrates the profile of a typical director at the centers surveyed.




                                        Page 52                                          GAO/HRDM-UBR   Early Childhood   Education
                          Section 4
                          What Were the Characteristics of and
                          Services Provided by the NAEYC
                          Centers Surveyed?




Figure 4.9


        GAO Profile of Directors in
                 NAEYC Centers
             l   On average directors:
                 aearn $24,430 per year
                 l have 15 years’ experience
             047%    have a graduate degree
                 @mostin early childhood
                 education




                           Page 53                               GAO/HRD9@-43BR   Early Childhood   Education
                                       Section 4
                                       What Were the Characteristics of and
                                       Services Provided by the NAJZYC
                                       Centers Surveyed?




Fiaure 4.10


       GAO Over Half of Teachers Had
           at Least Bachelor’s Degrees
                                                                 Some College
                                                                 2%
                                                                 Other



                                                                 High School Diplomas or Less

                                                                 Associate’s Degrees




                                                                 Bachelor’s Degrees or More
              Total teachersis 1,804




Teachers                               Of the teachers in the centers surveyed, 52 percent had at least bache-
                                       lor’s degrees (see fig. 4.10). Of these, a minimum of 38 percent had
                                       degrees in early childhood education or child development. An addi-
                                       tional 38 percent had associate’s degrees or some college education; of
                                       those with associate’s degrees, 90 percent had degrees in early child-
                                       hood education or child development. The remaining teachers had other
                                       training or high school diplomas or less. Figure 4.11 illustrates the pro-
                                       file of a typical teacher at the centers surveyed.




                                        Page 54                                        GAO/IIRlMO4BR   Early Childhood   Education
                          Section 4
                          What Were the Characteristics of and
                          Services Provided by the NAEYC
                          Centers Surveyed?




Figure 4.11


        GAO Profile of Teachers in
            NAEYC Centers

              l   On average teachers:
                  *earn $14,087 per year
                  *have 6 years experience
              l   52% have a 4-year college
                  degree or more
                  @manyin early childhood
                   education
                  038% of teachers have AA
                   degrees or some college




                          Page 55                                GAO/BRD-90-43BR   Early Childbood   Education
                                      Section 4
                                      What Were the Characteristics of and
                                      Services Provided by the NAJZYC
                                      Centers Surveyed?




Figure 4.12
r

        GAO Sixty Percent of Aides Had
            at Least Some College
                              r                                  EHighSchoo,




                                                                 High School Diplomas




                                                                 Associate’s Degrees


                                                                 Bachelor’s Degrees



                                                                 Some College
                Total TeacherAiies is 1,942




Teacher Aides                          Of all aides, 48 percent had associate’s degrees or some college training;
                                       12 percent had bachelor’s degrees (see fig. 4.12). Of the aides with bach-
                                       elor’s or associate’s degrees, 50 percent had degrees in early childhood
                                       education or child development. Figure 4.13 illustrates the profile of a
                                       typical teacher aide at the centers surveyed.




                                       Page 56                                         GAO/HRD90-43BR   Early chiwlood   Education
                           Section 4
                           What Were the Characteristics of and
                           Services Provided by the NAEYC
                           Centers Surveyed?




Figure 4.13


        GAO Profile of Teacher Aides in
                  NAEYC Centers
              l   On average aides:
                  *earn $10,219 per year
                  ahave 3 years’ experience
              l   92% have at least a high
                  school diploma
              022% have an AA degree or
               more




                          Page 57                                 GAO/HRD-9043BR   Early Childhood   Education
                                             Section 4
                                             What Were the Characteristics of and
                                             Services Provided by the NAEYC
                                             Centers Surveyed?




Figure 4.14


        G-0 Staff Experience in Early
            Childhood Education
                    Average Number of Years of Experfence




Experience Levels                            On average, directors of the early childhood education centers surveyed
                                             had 15 years of experience in the field; teachers, 6 years; and teacher
                                             aides, 3 years (see fig. 4.14).

                                              Teacher experience did not significantly vary by teacher education
                                              level, that is, teachers with bachelor’s degrees and those without such
                                              degrees had 6 years of experience. In addition, the teacher experience
                                              level was the same in urban and rural centers. The average years of
                                              experience was also the same across regions, except in the South, where
                                              teachers averaged 7 years of experience, about 1 year more than that of
                                              teachers in other regions.



                                              Page 68                               GAO/HRJHO48BR   Early Childhood   Education
                                         Section 4
                                         What Were the Characteristics of and
                                         Services Provided by the NAEYC
                                         Center8 Surveyed?




Figure 4.15


        GA!! Reasons for Centers’
             Staff Shortages
                 loo   Porwnf of contm

                  So

                  80

                  70
                  m-

                  so
                  so

                  50
                  50

                  40
                  40

                  so
                  so

                  20
                  20

                  10




                 Total centersis 79




                                         Similarly, teacher aides in the South averaged 4 years of experience,
                                         compared to an average of 3 years in other regions. In addition, teacher
                                         aides in both urban and rural centers had, on average, the same amount
                                         of experience, 3 years.

Turnover Rates                           The annual staff turnover rate was 26 percent for teachers and 54 per-
                                         cent for teacher aides. In addition, 79 centers reported operating with a
                                         staff shortage for 1 month or more in fiscal year 1988. As shown in
                                         figure 4.15, the primary causes underlying staff shortages were lack of
                                         qualified staff and low pay.




                                         Page 59                                GAO/HRD9@43BR   Early Childhood   Education
Appendix I

Comparisonof SelectedRequirements
for Early Childhood Education Programs:
S. 123 Criteria and NAEYC Standards
               Program
               requirement           S. 123 criteria                          NAEYC standards
               Curriculum            Developmentally appropriate for          Developmentally appropriate
                                     the child’s age and all areas of the     activrties and materials that are
                                     child’s development, rncludrng           selected to emphasize expenential
                                     educatronal, cognttive, physical,        learning
                                     emottonal, and socral
               Child-staff ratio     lot0 1                                   lot0   1
               for 4-year-olds
               Maxtmum group         20 children                              20 children
               size
               Staff
               qualifications
                  Teachers           State certtfication In early childhood   Associate degree in early childhood
                                     education or child development, if       education, child development, or
                                     available; nationally recognrzed         nationally recognized child
                                     chrld development credential; or         development credential
                                     signrfrcant college coursework In
                                     early chrldhood education
                  Aides/teacher      40 hours of preservice training          High school degree and preservice
                  assistants                                                  training
               In-service traintng   24 hours annually                        Required, but no minimum number
                                                                              of hours established
               Meals                 Must provide adequate and                Must ensure that children receive
                                     nutntious meals and, at parent’s         nutritious meals, but not required to
                                     request, breakfast                       provide them
               Supplementary         Must provide screentng for               Must provide health and social
               services              handicapping conditions and              service referrals and developmental
                                     health problems, information and         assessments of children
                                     referral services, and parent
                                     education
                                         --.__-
               Health and safety     Comply with applicable state and         Comply with applicable state and
                                     local laws and federal and state         local program requirements; staff
                                     standards                                trained to detect illness and at least
                                                                              one member trained in emergency
                                                                              aid




               Page 60                                              GAO/HRIMO43BR           Early Childhood   Education
Appendix II

Objectives,Scope,and Methodology


                      In February 1988, the Chairman, Senate Labor and Human Resources
                      Committee, requested that we obtain information on the costs of provid-
                      ing high-quality early childhood education to assist the Committee in its
                      ongoing deliberations over S. 123 (Smart Start: The Community Collabo-
                      rative for Early Childhood Development Act of 1989). In later discus-
                      sions with the Committee staff, we agreed to

              l estimate the average annual cost per child of providing high-quality
                early childhood education;
              . identify the specific costs of an early childhood education center’s
                budget;
              l compare the average annual salary for teachers of early childhood edu-
                cation with that for teachers in public elementary schools;
              9 determine the extent to which center costs change when certain factors,
                such as the number of children enrolled and the ratio of teaching staff to
                children, change; and
                identify the proportion of centers surveyed that were located in
                  l



                churches, synagogues, or other religious organizations.

                      To collect information on costs and services of programs of high-quality
                      early childhood education, we sent copies of a questionnaire to directors
                      of all of the 265 full-day, full-year preschool and early childhood educa-
                      tion programs accredited by the National Association for the Education
                      of Young Children,’ as of October 1988. As of October 1988, there were
                      658 accredited programs. (We did not survey those accredited programs
                      that operated only on a part-day or part-year basis or served only
                      school-aged children.) The 265 full-day programs surveyed may serve
                      children on a part-time as well as full-time basis.

                      NAEYC,  a membership organization of more than 70,000 professionals in
                      the field of child development and early childhood education, provides
                      the only national voluntary accreditation system exclusively for all
                      types of early childhood centers and schools. We surveyed NAEYC-
                      accredited programs because many of NAEYC’S accreditation standards
                      are similar to program criteria in S. 123 (see app. I). Thus, we believe
                      the costs of NAEYC-accredited programs would most likely be similar to
                      the costs of programs funded under the proposed bill. For example, both
                      S. 123 criteria and NAEYC’S standards require a maximum child-teacher


                      ‘The 265 excludes 20 of 28 centers that were part of eight programs that operated more than 1
                      center. In order that the administrators of these programs would not be burdened with completing a
                      questionnaire for more than one center, we asked that each of the eight administrators complete a
                      questionnaire for the largest center serving 4-yearolds in his or her program.



                      Page 61                                           GAO/HRIHO-43BR       Early Childhood   Education
Appendix II
Objectives, Scope, and Methodology




ratio of 10 to 1, a developmentally appropriate curriculum, and properly
trained staff.

Two nationally recognized experts in the area of economics and educa-
tion, W. Norton Grubb and Henry M. Levin, reviewed earlier drafts of
this report, and we have incorporated their comments throughout.
W. Norton Grubb is Professor of Education at the Graduate School of
Education of the University of California at Berkeley. Henry M. Levin is
Director of the Center for Educational Research at Stanford and Profes-
sor of Education and Economics at Stanford University.

Our questionnaire asked NAEYC-affik&d      centers to report information
on center costs, services, and children served for fiscal year 1988. Of the
265 centers in our survey, 78 percent (208) responded to our request for
information, The 265 centers represent the universe of full-day, full-
year NAEYC-accredited programs.

We visited five centers to (1) verify questionnaire responses and (2) test
the feasibility of respondents’ providing accurate data. We telephoned
all other respondents to follow up on their questionnaire responses, par-
ticularly those responses relating to salaries and other center costs.
Through center visits and telephone calls to respondents, we attempted
to minimize a potential for respondents to underreport center costs.

Our fiscal year 1988 regional estimates of 4-year-olds from families
below 115 percent of the poverty level were based on the 1980 Bureau
of the Census Survey. Using the Census Bureau’s March 1981 and March
1988 Current Population Surveys, we adjusted the data from the 1980
survey for population growth rate between 1980 and 1988.

We determined average annual center costs’ per child by region and
urban or rural location. For determining regional costs, we used the four
geographical regions- Northeast, Midwest, South, and West-desig-
nated by the Bureau of the Census. For determining costs by urban and
rural locations, we defined “urban” centers as those in counties that are




 2We used means--rather than medians-as measures of average costs, which resulted in slightly
 higher (more conservative) estimates of costs. Mean costs were generally less than 10 percent higher
 than median costs.



 Page 82                                            GAO/HRD-9043BR        Early Childhood   Education
Appendix II
Objectives, Scope, and Methodology




in metropolitan statistical areas3 and “rural”                     centers as those located in
nonmetropolitan statistical areas.

We believe that information obtained from the centers provides reliable
data for developing a reasonable estimate of the cost per child of provid-
ing full-day, full-year high-quality early childhood education.” Yet these
centers-which      sought and obtained accreditation on a voluntary
basis-may not be statistically representative of all high-quality early
childhood education centers in the nation. Therefore, our report data are
not necessarily representative of costs nationwide.

For example, the nation’s five largest chains of child care providers,”
representing almost 2,700 child care centers, generally are not NAEYC-
accredited, but they are state-licensed. According to officials at four of
the five chains,” the chains require centers, at a minimum, to meet state
licensing standards. Thus, program criteria, such as those regarding
maximum child-to-staff ratios or group sizes, can be expected to vary
from state to state.

Although all centers we surveyed enrolled 4-year-olds, some centers also
enrolled children of other ages. For example, 33 percent of the centers
served infants and 45 percent served toddlers. Lower child-staff ratios
necessary for serving infants and toddlers generally contribute to higher
per child costs at these centers. Although our data suggest that centers
not serving infants and toddlers have lower costs than centers that do,
our data do not allow us to determine the cost difference between cen-
ters serving infants and toddlers and those not serving such children.

We obtained and reported information on the documented benefits of
several early childhood education programs. However, our cost data are
not necessarily representative of the costs of these programs, some of
which were experimental and had lower child-staff ratios than the
XAEYC centers we surveyed.


 :rMetropolitan statistical areas are defied by the Office of Management and Budget as having one or
 more central counties with an urbanized area of at least 50,000 inhabitants. Metropolitan statistical
 areas may also include outlying counties that have close economic and social ties with the central
 counties.

 ‘These data do not reflect, and should not be used to estimate, the cost of part-day or part-year
 programs.

 “These include KinderCare. La Petit Academy, Children’s World Learning Centers, Gerber’s Chil-
 dren’s Centers, and Children’s Discovery Centers.
 “‘Officials at one of the five chains did not respond to our inquiry



 Page 63                                                GAO/BRD9043BR       Early Childhood   Education
Appendix II
Objectives, Scope, and Methodology




We based our calculation of average cost per child on centers’ reported
enrollment of full-time and part-time children. We determined that for
35 centers in our survey, on average, part-time children attended their
centers 42 percent of the time that full-time children attended. We,
therefore, weighted part-time children accordingly when calculating the
cost per child.

Many centers we surveyed indicated that they received in-kind dona-
tions, which reduced outlays they would otherwise have had to make
for such items as rent and repairs, equipment and materials, and supple-
mentary services. Therefore, we included donations received by centers
in calculating the average cost per child. We then added to the center’s
reported cost per child the value of the donations as estimated by the
center. Assuming that in-kind donations will continue to be obtained by
centers to the same degree in the future as they have been in the past,
our estimate of average annual cost per child (which includes the value
of donated services) overestimates the actual costs that will be borne
directly by centers.

In addition, many centers reported no costs for (1) rent or mortgage or
(2) insurance-costs   necessary for operating an early childhood educa-
tion center-but did not indicate whether these costs were donated or, if
donations were received, the value of the donations. To correct for a
possible underestimation of costs for these centers, we estimated
(1) rent and mortgage and (2) insurance costs by determining the aver-
age in each geographic region for those centers reporting such costs.

For example, a center that occupies space which it owns and for which
the mortgage has been paid off might have reported no expenditures for
occupancy. The cost of the space, in this case, is the rent forgone that
could have been earned if the space was rented to others. Again, as in
the case of in-kind donations, our estimate of cost per child, which
includes the estimated value of the space and insurance coverage, over-
states the costs that will be borne directly by centers.

We compared early childhood education teacher salaries with those of
public elementary school teachers by region (see p. 27).’ To obtain com-
parable data on public school elementary teacher salaries, we used esti-
mates from Estimates of School Statistics: 1987-88, collected from states

7For the purpose of this comparison, we used only salary data reported for early childhood education
teachers working 35 or more hours per week. The vast majority of public school teachers worked full
time.



 Page 64                                           GAO/HRD9643BR        Early Childhood   Education
Appendix 11
Objectives, Scope, and Methodology




by the National Education Association. According to the association’s
manager of research services, these estimates are determined by divid-
ing the total salaries for elementary school teachers by the number of
public elementary school teachers for each state. To determine an aver-
age salary for the nation as a whole and for the regions in which these
states are located, we in turn weighted state averages by the numbers of
those teachers in each state.

In addition, we compared the average salaries of those full-time teachers
of early childhood education who had bachelor’s degrees and various
experience with estimates of average salaries of full-time public and pri-
vate school teachers with similar experience. We included private school
teacher salaries in the comparison because a majority of the early child-
hood education centers we surveyed are privately operated. We included
both elementary and secondary school teachers in the comparison
because salary data were not available separately for private school ele-
mentary teachers by years of experience. We included in the comparison
only teachers of early childhood education with bachelor’s degrees since
more than 95 percent of all public and private school teachers have such
degrees.

We obtained data on roughly 8,300 public school teachers and 4,700 pri-
vate school teachers from surveys conducted by the Department of Edu-
cation’s National Center for Education Statistics. Because the data on
public and private school teachers reported by NCES were for school year
1985-86, we adjusted the salary levels to those of 1987-88, so that they
would be in line with fiscal year 1988 salaries reported by the early
childhood education centers we surveyed. From school year 1985-86 to
1986-87, we used an inflation factor of 5.4; from school year 1986-87 to
1987-88, we used a factor of 5.5.

Most of the early childhood education teachers in our survey worked in
centers for 12 months of the year. We annualized salaries for those
teachers who worked less than 12 months. On the other hand, the public
and private school teacher salaries, as reported by KCES,are based on a
school year that we assumed to be about 10 months. Therefore, we made
and reported two separate comparisons of salaries of early childhood
education teachers with those of public and private school teachers. In
one comparison, public and private school teachers’ salaries are based
on a school year of 10 months; and in the other, such salaries are
annualized.




 Page 65                              GAO/HRB9043BR   Early Childhood   Education
Appendix II
Objectives, Scope, and Methodology




To determine center cost for insurance, we averaged the costs of the 168
centers in our survey that reported such costs. The other 40 centers
reported no costs for insurance. Presumably, their insurance costs were
paid by their program sponsors, for example, churches, public school
systems, or hospitals. We determined the average center cost for each of
the four supplementary services required under S. 123 and for each of
the three other supplementary services not required. To determine the
center cost for each supplementary service, we averaged the costs of the
centers in our survey that reported such costs or indicated the value of
in-kind donations received for that service.

In identifying which centers surveyed were religiously affiliated, we
considered a “religiously affiliated” center to be one in which a religious
organization is involved in the center’s administration or has control
over the content or structure of the program, the hiring of personnel, or
the selection of children.

We used regression analysis to estimate the impact of various factors on
the total cost of operating an early childhood education center. The
analysis was conducted in two parts: (1) a total cost equation to estimate
the direct influences on total cost and (2) wage equations (one for teach-
ers and one for aides) to estimate the indirect influences, that is, factors
that affect total cost by affecting wages of teachers or aides.RA total of
187 centers with usable data were included in the analysis.




‘For example, centers that only hire teachers with many years of experience must pay higher wages
to attract qualified workers. Thus, because the experience requirement increases salary costs, it indi-
rectly raises total costs. Other factors, such as the number of full-time equivalent children, directly
affect total cost.

The cost of operating a center was hypothesized to depend on the wages paid to the staff (directors.
teachers, and aides); the rental, mortgage, and maintenance costs of the center; the cost of other
supplies; the number of full-time equivalent children; and several factors thought to be associated
with center quality (such as average group size and child-staff ratio), as well as location and the
percentage of children from low-income families served by the center.

The hourly wage of teachers and aides was hypothesized to be affected by their education and
experience, full-time or part-time status, benefits received (such as paid vacation and paid health
benefits), working conditions and requirements (such as child-staff ratio and percentage of children
with handicapping conditions), and other center-specific factors (such as for-profit or nonprofit spon-
sorship, total number of adults employed, as well as regional and urban or rural location).



Page 66                                              GAO/HRD9@43BR        Early Childhood    Education
Appendix III

TablesSupporting Figures in Report Text


Table 111.1:National Cost Per Child
(Data for Fig. 2.1)                                                                 costs
                                                                             repo{;g;~
                                                                                                     In-kind                        Total
                                                                                 centers          donations            Total respondents
                                        Nation                                     $4,200               $597          $4,797               205
                                        Urban                                       4,339                605           4,944               181
                                        Rural                                       3,154                536           3,690                24


Table 111.2:Regional Cost Per Child
(Data for Fig. 2.2)                                                 costs                                         Donations
                                                            wo;;$
                                                                                  In-kind                        percenta::          Total
                                                                centers        donations                Total        of total resoondents
                                        Northeast                   $4,949               $659          $5,608              12                 28
                                        Midwest                      4,286                465           4,751              10                 73
                                        South                        3,978                711           4,689              15                 67
                                        West                         3664                 608           4,472              14                 37


Table 111.3:Cost Per Child by Center
Sponsor (Data for Fig. 2.3)                                                         costs
                                                                             wo;;g$g
                                                                                                      In-kind                        Total
                                                                                  centers          donations            Total respondents
                                        Public                                      $4,467               $987          $5,454                 22
                                        Private nonprofit                            4,211                614           4,025                 153
                                        For-profit                                   3,947                226           4,173                  30


Table 111.4:Cost Per Child at Centers
Sewing Children of Different Income                                                 costs
Levels (Data for Fig. 2.4)                                                   woG;g;g
                                                                                                      In-kind                        Total
                                        Center target group                       centers          donations            Total respondents
                                        Serves no low-income children               $4,752               $731          $5,403                  43
                                        Serves >O but ~25%                           4,034                533           4,567                  69
                                        Serves ~25% low-income
                                        children                                       4,263              687           4,950                  63


Table Ill.5 Teacher Salaries in Urban
Compared With Rural Areas                                                                                          Teacher                Total
(Data for Fig. 2.8)                                                                                                salaries     reSDOndentS
                                        Nation                                                                      $14,087               1,137
                                        Urban                                                                        14,400               1 030
                                        Rural                                                                        11.056                 107




                                        Page 67                                                 GAO/BRD-9043BR     Early Childhood   Education
                                             Appendix III
                                             Tables Supporting     Figures in Report Text




Table 111.6:Salaries of Earlv Education Teachers and Public and Private School Teachers by Years of Experience (Data for Fig. 2.9)
                                            Early childhood                      Public                     Private
                                                  education         Total      teacher         Total       teacher            Total
Years of experience                        teacher salaries respondents        salaries respondents        salaries respondents
5 years or less                                       $14,460               306         $19,453              1,068      $13,556             1,409
6to 10 years                                           15,324               117          23,015              1,586       16,124             1,126
11 to15years                                           17,582        __-     29          27,381              1,933       18,253               830
16to 20 years                                          16,982                21          30,801              1,570       20,092               461
21 to25 Years                                          18,448                 9          32,117                922       18,399               277




Table 111.7:Salaries of Early Education Teachers and Public and Private School Teachers by Years of Experience (Annualized) (Data
for Fig. 2 10)
                                            Early childhood                              Public                         Private
                                                  education               Total        teacher         Total           teacher         Total
Years of experience                        teacher salaries        respondents         salaries respondents            salaries respondents
5years or less                                        $14,460               306         $23,452               1,068     $16,267             1,409
6to 10 years                                           15,324               117          27,617               1,586      19,348             1,126
11 to 15years                                          17,582                29          32,857               1,933      21,904               830
16to20years                                            16,982                21          36,961               1,570      24,111               461
21 to25 Years                                          18,448                 9          38,540                 922      22,078               277




Table 111.6:Average Directors’ and Teacher Aides’ Salaries by Region (Data for Figs. 2.7 and 2.11)
                                       Total                       Total                           Total                               Total
                    Northeast respondents         Midwest respondents               West respondents                      South respondents
Directors              $25,041             32         $25,007                 74            $21,462              37     $24,853                 71
Teacher aldes           10,213             97             10,407             255              9,244             137       10,639              206




                                                Page 68                                               GA0/HRD9043BR   Early Childhood   Education
                                               Appendix III
                                               Tables Supporting   Figures in Report Text




Table 111.9:Centers Offering Certain
Benefits to All or Some leachers                                         All teachers                 Some teachers
(Data for Fig. 2.12)                                                              Percentage                  Percentage           Total
                                               Benefit                 Centers        of total       Centers       of total respondents
                                               Health
                                               insurance                    136             65             33              16               208
                                               Pension and
                                               retirement
                                               coverage                      77             37              16              8               208
                                               Life Insurance                85             41              17              8               208
                                               Reduced child
                                               care fees                    101             49              11              5               208
                                               Vacation leave
                                               (paid)
                                               ~1       ,
                                                                            180             87              19              9               208
                                               Sick leave
                                               (paid)                        183            88              14               7              208


Table 111.10:Centers in Public and Private
Sectors (Data for Fig. 4.2)                                                                                                         Total
                                               Center sector                                          Number         Percent respondents
                                               Private for-orofit                                           31             15               208
                                               Public school-based                                           3              1               208
                                               Other public                                                 19              9               208
                                               Private nonprofit                                           155             75               208


Table 111.11:Center Size (Data for Fig. 4.3)
                                                                                                                                    Total
                                                                                                      Number         Percent respondents
                                               Small                                                        75             36               208
                                               Medium                                                      106             51               208
                                               Large                                                        24             12               208
                                               Very large                                                    3              1               208


Table 111.12:Median Monthly Fees for
Families by Income (Data for Fig. 4.5)                                                                           Median fees
                                                Income                                             Family of 2    Family of 4        Family of 6
                                               $8,000                                                      $81            $34               $26
                                               15,000                                                      180            121                90
                                               35,000                                                      280            200               170
                                               40.000                                                      280            280               280
                                               75,000                                                      280            280               280




                                                    Page 69                                      GAO/IiRLHM3BR     Early Childhood    Education
                                           Appendix III
                                           Tables Supporting    Figures   in Report Text




Table 111.13:Enrollments by Age Group
(Data for Fig. 4.6)                                                                                        Percentage            Total
                                           Ape croup of enrolled children                      Centers      of centers    respondents
                                           Infants                                                  69              33              208
                                           Toddlers                                                 94              45              208
                                           2-year-olds                                             169              81              208
                                           3-year-olds                                             207              99              208
                                           4-vear-olds                                             207              99              208
                                           5-year-olds                                             198              95              208
                                           Older than 5 years                                      114              55              208


Table 111.14:Staff Experience in Early
Childhood Education (Data for Fig. 4.14)                                                                      Years of       Total
                                           Staff                                                           experience respondents
                                           Directors                                                               14.6              207
                                           Teachers                                                                 6.3            1.739
                                           Teacher aides                                                            3.3            1,625


Table Ill.15 Reasons for Centers’ Staff
Shortages (Data for Fig. 4.15)                                                                                                Total
                                                                                               Number          Percent respondents
                                            Unable to find qualified staff                           55              70               79
                                            Staff would not work for pay level offered               50              63               79
                                            Lack of center resources to hire new staff                8              10               79
                                            Other                                                    12              15               79




                                            Page 70                                        GAO/HRD!W43BR     Early Childhood   Education
Technical Description of GAO’s Economic
Analysis of the Costsof High-Quality Early
Childhood Education
                         To learn how various factors affect the cost of operating an early child-
                         hood education center, we developed two related economic models. The
                         first model examines the direct influences of input prices, center charac-
                         teristics, and location on center cost. The second model analyzes those
                         factors that affect the wages of teachers and aides. Because labor is an
                         important input, any change that affects wages will have a substantial
                         (indirect) impact on the cost of early childhood education. By analyzing
                         the results of both models, we were able to understand how various fac-
                         tors directly and indirectly affect center cost.

                         We estimated both models using multiple regression-a standard statis-
                         tical technique that quantifies the relationship between a dependent
                         variable and a set of independent variables. The construction of each
                         model (center cost and wage) and its results are discussed below.

                         Our cost model is derived strictly from economic theory. We assume that
Center Cost Regression   all centers attempt to minimize total cost for any given center size and
Model                    quality level.’ The total annual center cost is hypothesized to depend on
                         the price of inputs used in the production process, the amount of output
                         produced, and characteristics of the center. We used multiple regression
                         to estimate the parameters of the cost equation and quantify the rela-
                         tionship between total annual cost, input prices, level of output, and




                         ‘All centers are assumed to be efficient (in the sense that they minimize the cost of producing any
                         given level of output) because penalties exist for inefficiency. In a competitive environment, ineffi-
                         cient firms are soon driven out of business. In a nonprofit environment-where the vast majority of
                         the early childhood education centers operate-the consequence of inefficiency is not bankruptcy,
                         but a reduction in the number of children that can be served with a given amount of funds. Thus, it is
                         reasonable to believe that even nonprofit centers attempt to be efficient.



                         Page 71                                             GAO/HRD9043BR        Early Childhood   Education
Appendix IV
Technical Description  of GAO’s Economic
Analysis of the Costs of High-Quality Early
Childhood Education




center characteristics2 A complete definition for each variable is con-
tained in table IV. 1.

Total cost is the total annual cost of operating a center. It is calculated
as the sum of all explicit (out-of-pocket) costs plus the reported value of
all donated labor, supplies, equipment, services, and space.

For each center we calculated five input price variables: the average
hourly wage rate of teachers (WAGET) and aides (WAGEA); the average
monthly wage of the directors3 (WAGED); the occupancy cost-annual
rent or mortgage payments plus repair and maintenance costs-per
square foot of total center space (RENT); and the combined cost of sup-
plies, equipment, supplementary services, insurance, and nonteaching
labor-all divided by the number of FTE children (OCOST). Many cen-
ters reported receiving donations in one or more of the above categories.
In those instances, the self-reported value of the donation was included
in the computation of the input price.4




2We chose the Cobb-Douglas functional form for the cost equation. This functional form has been
widely employed in economic research and fulfills the economic theoretical requirements for a cost
function.

Because output quality is not measured and may vary between centers, we added an additional cate
gory to the standard cost function that contains center quality indicators as well as measures of other
center characteristics. The characteristics in this category (denoted by Q) are summarised by a non-
negative index function:

                             Q = exp(Z’B),
where Z is a vector of center characteristics and B a vector of weights (to be estimated in the cost
equation).
Consequently, the estimated cost equation can be expressed as:
                             InC = f(mP, lnY, Z),

where C is total center cost, Y is output, P is a vector of input prices, and Z is defined as above.

“Includes assistant directors.

‘For example, many centers reported that the monthly rent they paid was below market value. For
those centers we substituted the fair market value (as determined and reported by the center) for the
rental payment. The difference between the fair market value and the annual amount actually paid is
considered to be the value of donated space.



Page 72                                                GAO/HBD-9043BR        Early Childhood     Education
Appendix IV
Technical Description  of GAO’s Economic
Analysis of the Costa of High-Quality Early
Childhood Education




We measure the output of a center by the number of FTE children
(CHILDREX) enrolled at that center. It should be noted that our mea-
sure represents a proxy, albeit the best available one, for the true out-
put “education” that each center produces.” Because higher quality
centers likely face higher operating costs than lower quality centers, it is
important to control for any quality differentials that may exist to pre-
vent included variables--especially CHILDREN-from         serving as a
proxy for quality.” However, as the true “value added” output is unob-
served, intrinsic center quality can only be inferred from center and
input characteristics. Our model includes two observed center character-
istics-average children’s group size (GROUPSZ) and the child to adult
ratio (CARATIO)-thought      to be partial indicators of center quality.
Everything else equal, centers with smaller children’s group sizes and
lower child/adult ratios are believed to be superior. Centers may
enhance quality by improving the quality of the inputs used-hiring
teachers with more education or aides with more years of experience.
These latter types of quality differentials are reflected in the price paid
for a given input and affect total cost indirectly through higher input
prices.

Finally, we included several other variables that could help explain cost
differentials between centers. OUTPC measures the amount of outdoor
space in square feet (in thousands) per child. Centers with large
amounts of outdoor space-some have as much as 10 acres-likely
have higher costs compared to other centers in similar areas. Because
recently established centers are likely to be smaller than older ones, the
dummy variable NEW is included to prevent the center output variable,
CHILDREX, from picking up any cost differences due to center age.
NEW is equal to one if the center has been in operation for less than 5
years and zero otherwise. PROFIT and INFANTS are also binary vari-
ables. PROFIT indicates that the center is for-profit, while INFANTS
indicates that the center serves children under 2 years of age. For-profit
centers may have lower costs than nonprofit centers if the forces of
competition are needed to insure economic efficiency. Centers that serve
infants, in addition to 4-year-olds, are likely to have higher costs than
centers that serve only 4-year-olds. Center cost is hypothesized to


“A true output measure could only be constructed if we observed the “value added-‘-that is the
increase in knowledge, skrlls, and capabilities (all broadly definedhimparted in each child and, in
addition, could quantify that (multidimensional) output in some meaningful way.

“All of the early childhood centers included in our analysis are accredited by NAEYC. As such, they
all meet minimum quality standards determined by NAEYC. Nonetheless, within this group, quality
levels may vary.



 Page 73                                            GAO/HRD-!M-43BR      Early Childhood   Education
                       Appendix N
                       Technical Description  of GAO’s Economic
                       Analysis of the Costs of High-Quality Early
                       Childhood Education




                       increase with the percentage of children that are handicapped (HANDI-
                       CAP).7 Finally, two variables- location in a metropolitan statistical arei
                       (URBAN) and the percentage of children from low-income families
                       @INCOME)-are        included to capture any cost differentials due to
                       center location.


Cost Model Empirical   We estimated the regression models by the method of ordinary least
                       squares for 187 centers with usable data. Table IV.2 presents the esti-
Results                mates of the regression coefficients, the standard error for each of the
                       estimated coefficients, and the t-statistic for the null hypothesis that the
                       true parameter value is equal to zero. Because the dependent variable is
                       measured in logarithms, the estimated coefficients show the percentage
                       change in a center’s total cost caused by a l-percent change in CHIL-
                       DREN or any of the input price variables. For a one-unit change in any
                       of the other independent variables, the estimated coefficients show the
                       approximate percentage change in total cost.

                       An estimate is considered statistically significant if the probability is
                       low that the true value of the coefficient is zero. We chose as our crite-
                       rion a significance level of 0.05; that is, we required that the probability
                       of the true coefficient being zero is no greater than 0.05. The critical
                       t-statistic (two-tailed test), given the size of our data set, is approxi-
                       mately 1.96. The number of children, child-to-adult ratio, and all of the
                       input price variables have a significant effect on center cost. Two addi-
                       tional variables introduced mainly as control variables, the amount of
                       outdoor space per child and an indicator for whether the center serves
                       infants, are also statistically significant.

                       Our results indicate that statistically significant economies of scale exist
                       in early childhood education, that is, average cost (cost per child) falls
                       as center size increases. The estimated coefficient on CHILDREN indi-
                       cates that a lo-percent increase in center size (measured by the number
                       of full-time-equivalent (FTE) children) would decrease average cost by
                       about 2 percent.X Economies of scale existed even when we restricted



                       7HANDICAP is measured as the proportion of children in each center that were emotionally dis-
                       turbed; mentally retarded; developmentally delayed; speech, hearing, visually, or orthopedically
                       impaired; deaf and blind; multihandicapped; or otherwise health impaired.

                        ‘Economies of scale are said to exist because the estimated coefficient on CHILDREN (0.8) is statrsti-
                        tally less than 1.Oat the 1-percent level of significance. A lo-percent increase in center size causes
                        total cost to rise by only 8 percent; thus average cost decreases by 2 percent.



                        Page 74                                             GAO/HRLWO43BR         Early Childhood   Education
Appendix Iv
Technical Description  of GAO’s Economic
Analysis of the Costa of High-Quality Early
Childhood Education




our analysis to include only centers with more than 68 FTE children.9 It
is unlikely that the cost advantage of large centers is due to their age
and efficiencies gained from experience. We found no cost differential
between centers that were less than 5 years old and those that had been
in business longer.

The findings of scale economies must be interpreted with care, as the
implied cost advantage of large centers may be overstated for two rea-
sons. First, if the administrative burden on each director increases with
center size, then larger centers might have to offer higher salaries than
those paid at smaller centers to attract capable administrators. Higher
salaries would at least partially offset the cost per child differential
between small and large centers. Second, an increase in’center size
achieved by the consolidation of many small centers into fewer larger
centers would impose additional costs on parents. The decrease in the
number of centers would increase the average commuting distance for
each child and thus increase commuting costs for parents.lO Thus center
size may be limited by local conditions-such     as population density-
that affect the number of children potentially served by any given
center-l*

Total costs rise significantly with all five of the measured input prices
(WAGET, WAGEA, WAGED, RENT, and OCOST). Of the three variables
that represent labor costs, total cost is most sensitive to changes in the
wages of teachers. A lo-percent increase in teachers’ wages increases



‘Because the Cobb-Douglas imposes constant elasticity of cost with respect to output, we initially
divided the sample into two groups-centers with no more than 68 FTE children and those with more
than 68 FTE children--and estimated the cost equation separately for the two groups. However, all
centers were pooled together when an F-test failed to reject the null hypothesis that the estimated
coefficients in the first group were identical to the coefficients in the second group.
Although not reported here, we also estimated a quadratic average cost equation. Those results indi-
cate that scale economies are eventually exhausted: the lowest cost per child occurred at a center size
of 287. Only 2 out of 208 centers reported serving more than 287 FlX children.

‘°Commuting costs include both direct costs (for example, costs of using public transportation or
operating a private vehicle) and time costs (the value that parents place on additional minutes of
commuting time).

“A rural center-because of the sparseness of the surrounding population-would have to attract
children from farther away than an urban center of the same size. This means that the average direct
transportation costs would be higher for rural parents than for urban parents. However, if wages are
lower in rural areas, the time or opportunity costs would be less than in urban areas. Thus it is
unclear whether rural or urban centers are best able to lower the cost per child by increasing center
size. In our sample of 208 centers, the average number of Fl’E children was nearly identical for rural
and urban centers (83 and 81, respectively).



Page 76                                              GAO/HRD9O4BR         Early Childhood   Education
                    Appendix IV
                    Technical Description  of GAO’s Economic
                    Analysis of the Costa of High-Quality Early
                    Childhood Education




                    total cost by 3 percent. In contrast, the same increase in aides’ and direc-
                    tors’ wages would increase total cost by only 1.3 and 1.7 percent,
                    respectively.

                    Of the two variables introduced to capture quality differences between
                    centers directly, only one, the child-to-staff ratio, had a statistically sig-
                    nificant impact on total cost. An increase of one child per staff, holding
                    all other factors constant, reduced costs by 4.4 percent.


                    Using economic theory, we developed a model of wage determination for
Economic Model of   teachers and aides in the early childhood education industry. The wage
WageDetermination   equation model relates the wage rates to factors-specific   to the indi-
                    vidual worker, the center where the worker is employed, or the location
                    of the center-hypothesized    to influence wages.12We then used the tech-
                    nique of ordinary least squares to obtain estimates of the coefficients in
                    our multiple regression model. Although the basic model is identical for
                    both teachers and aides, the impact of specific factors may be dissimilar
                    for the two groups. Thus, teachers and aides were analyzed separately.

                    According to economic theory, wages of workers are determined in the
                    labor market through the interaction of the supply of labor by workers
                    and the demand for labor by firms. The supply of workers is determined
                    by the level of skills or knowledge required in an occupation and the
                    economic opportunities offered in alternative industries. In general, the
                    demand for workers is a “derived demand,” that is, a firm’s demand for
                    workers is determined by the demand for the good or service produced
                    by the firm, as well as the availability and relative prices of substitute
                    inputs. Thus, factors that influence the demand for early childhood edu-
                    cation will also affect the demand for teachers and aides.

                    Each of the variables in our model originate from one of three basic cat-
                    egories suggested by economic theory:

                     1. Measures of a worker’s human capital, i.e., skills and training.

                     2. Working conditions and nonpay compensation of the job.

                     3. Factors that influence demand.

                     ’ “Because our model includes information on centers as well as teachers and aides, it is considered to
                     be a “reduced-form” wage equation-that is, one that incorporates both demand for labor and supply
                     of labor effects.



                     Page 76                                              GAO/BRIM@43BR        Early Childhood   Education
Appendix IV
Technical Description  of GAO’s Economic
Analysis of the Costs of HighQuality Early
Childhood Education




The level of a worker’s “human capital” is captured by the number of
years of formal education (EDUC) and years of employment experience
(EXP) in early childhood education or childhood development. We
expect that wages will increase with both education and experience.
However, because additional years of experience may not increase
wages at a constant rate, EXP is entered quadratically in the wage
equation.

A number of variables were introduced to control for working conditions
and nonpay compensation. These variables are defined for each teacher
or aide, based on the center where the worker is employed. Working
conditions include the child-to-adult ratio13 (CARATIO) and the percent-
age of children that are handicapped (HANDICAP). Nonpay compensa-
tion factors are captured by three dummy variables: LEAVE, REDFEE,
and HEALTH. Each of these variables indicates if the teacher/aide
works at a center that provides some or all of its teachers/aides with the
specified fringe benefit: paid leave or vacation time (LEAVE), reduced
fees for care of employees’ children (REDFEE), and fully or partially
paid health insurance (HEALTH).

Because the wage equation incorporates both demand-side and supply-
side effects, the expected effect of CARAT10 is ambiguous; a higher
child-to-adult ratio could increase or decrease wages. That is, a lower
child-to-adult ratio indicates easier working conditions and should
result-other    things being equal-in workers being willing to work for
lower wages. However, if centers with a low CARAT10 (a possible indi-
cation of center quality) employ only the best teachers, and years of
education and experience do not completely control for teacher quality,
then a positive correlation between CARAT10 and WAGE would exist.

 The expected effect of HANDICAP and the three nonpay compensation
 variables on wages is unambiguous. If the percentage of handicapped
 children directly influences the difficulty of the job, wages should rise
 along with HANDICAP. The existence of any of the three fringe benefits
 is expected to lower the wage received because some of the worker’s
 compensation is received in a nonmonetary form.

 The percentage of children from low-income families (U)INCOME) is
 expected to have a negative effect on wages for two reasons. As the
 percentage of children from low-income families increases, the demand


 ‘?he number of adults is calculated as the total number of teachers, aides, and directors.



 Page 77                                             GA0/HRD9043BR        Early Childhood     Education
                    Appendix IV
                    Technical Description  of GAO’s Economic
                    Analysis of the Costs of BighQmlity  Early
                    Childhood Education




                    for early childhood education- and hence teachers-may fall. In addi-
                    tion, this variable may serve as a proxy for local price levels-low
                    wages and low prices may be seen typically in areas with many low-
                    income families .

                    Centers that are for-profit (PROFIT) or are larger in terms of the total
                    number of employees (ADULTS-total      number of teachers, aides, and
                    directors) may not pay the same wage rate as nonprofit or smaller
                    centers.

                    Location factors capture the net influence on wages from both demand
                    and supply influences that are not explicitly represented by the other
                    variables. For example, teachers who work in urban areas (URBAN) art
                    likely to command higher wages than teachers in rural areas both
                    because of superior alternative job opportunities in urban areas (a sup-
                    ply of labor effect), and because the demand for early childhood educa-
                    tion is greater (a demand for labor effect). Because differences in supply
                    and/or demand may exist between regions, we also introduced a set of
                    dummy variables that controls for teachers’ and aides’ regional location
                    (MIDWEST, WEST, SOUTH, NORTHEAST).


WageModel Results   We estimated the regression models by the method of ordinary least
                    squares. This was done separately for the 1,280 teachers and 1,423
                    aides. Table IV.3 presents the estimates of the regression coefficients
                    from the two wage equations. Because the dependent variable is mea-
                    sured in logarithms, the estimated coefficients show the percentage
                    change in the wage rate caused by a one-unit change in the independent
                    variable. The table also reports the standard error for each of the esti-
                    mated coefficients and the t-statistic for the null hypothesis that the
                    true parameter value is equal to zero.

                    As in the center cost regression, we chose as our criterion a significance
                    level of 0.05. The critical t-statistic (two-tailed test), given the size of our
                    data set, is approximately 1.96. Almost all of the estimated coefficients
                    are of the expected sign and significant at the 5-percent level or better.

                    An additional year of education increases wages by 6.0 percent for
                    teachers and 3.3 percent for aides. Additional years of experience
                    increase wages, but at a decreasing rate for both teachers and aides. For
                    example, our results indicate that an additional year of experience
                    would increase the wage rate by 2.5 percent for a teacher with 5 years
                    of experience, but by only 1.6 percent for a teacher with 10 years of


                     Page 78                                     GAO/HRIMO4BR   Early Childhood   Education
Appendix IV
Technical Description  of GAO’s Economic
Analysis of the Costs of B&Quality   Early
Childhood Education




experience. A similar result held for aides. An additional year of experi-
ence would increase by 1.6 percent the wage of an aide with 5 years of
experience, but would increase by only 1.0 percent the wage of an aide
with 10 years of experience.

The wages of both teachers and aides are higher in centers that enroll a
larger proportion of handicapped children. Our results indicate that
each percentage point increase in the number of handicapped children
results in a 0.6-percent increase in teachers’ wages and a 0.2-percent
increase in aides’ wages. The child-to-adult ratio had no statistical
impact on wages. The provision of fringe benefits as compensation fac-
tors are also important determinants of wages. As expected, centers that
provided paid vacation time offered lower wages-19.5 percent lower
for teachers and 11.5 percent lower for aides-than centers without a
paid leave policy. Centers that offered reduced child care fees for chil-
dren of employees paid lower wages to teachers (10.5 percent), but we
found no statistical impact on aides’ wages. Surprisingly, there was a
statistically significant positive relationship between wages and a
center’s provision (full or partial) of health insurance. Centers with
health insurance plans paid 11.5 percent higher wages to teachers and
 15.1 percent higher wages to aides.‘”

Other variables that controlled for the size, profit-making status, and
location of the center were found to be important determinants of
wages. Wages were higher for teachers, but lower for aides in centers
that were larger in terms of employment levels, although the magnitude
of the difference was relatively small.ls Relative to nonprofit centers,
for-profit centers paid 7.2 percent lower wages to aides; there was no
statistical difference in the teachers’ wages. Urban centers paid wages
that were 19.8 and 9.3 percent higher for teachers and aides, respec-
tively. The set of regional dummy variables was statistically significant.
Wages were lower in the West, Midwest, and South than in the North-
east. The proportion of low-income children in a center also affected
wages, lowering them for both teachers and aides. Wages of full-time

“Our measures of fringe benefits are imperfect. We do not know if an individual teacher received the
benefit; instead we only know if the benefit was offered to all or some of the teachers or some or all
of the aides working at a particular center. In addition, we do not know the value of the benefit
received. This shortcoming is especially relevant for health insurance as there may be great variation
in the proportion of this benefit that is paid by the center. Thus our HEALTH variable may be serv-
ing, in part, as a proxy for an omitted indicator of average center employee quality, rather than as a
direct measure of nonwage compensation--subsidized health insurance.

“The average center employed a total of 18 teachers, aides, and directors. Our results indicate that a
center with 10 additional employees would pay wages 1.6 percent higher for teachers and 1.0 percent
lower for aides.



Page 79                                             GAO/BRD9O43BR        Early Childhood   Education
                                   Appendix N
                                   Technical Description  of GAO’s Economic
                                   Analysis of the Costs of HighQuaLity Early
                                   Childhood Education




                                   teachers were 7.6 percent higher than wages of part-time teachers. Full
                                   time aides received 2.6 percent higher wages than part-time aides.

Table IV.l: Variable Definitions
                                   Variable                  Definition
                                   WAGET          -_         Average hourly wage, teacher
                                   WAGEA                     Average hourly wage, arde
                                   WAGED                     Monthly wage, director
                                   RENT                      Total annual rent or mortgage cost (including the value of donated
                                                             or subsidized space) plus annual matntenance and repair costs,
                                                             divided by the total number of square feet of space
                                   OCOST                     Cost of supplies and other miscellaneous costs, divided by the
                                                             number of FTE children
                                   CHILDREN                  Number of FTE chrldren
                                   CARAT10                   Child-to-adult ratio-number   of FTE children divided by the total
                                                             number of directors, teachers, and aides
                                   GROUPSZ             ---   Average children group size
                                   OUTPC                     Number of square feet (in thousands) of outdoor space per FTE
                                                             child
                                   NEW                       Equals 1 if center has been In extstence for less than 5 years, 0
                                                             otherwise
                                   PROFIT                    Equals 1 if center is for-profit, 0 otherwise
                                   HANDICAP             .__- coportron of handicapped chtldren
                                   INFANTS                   Equals 1 if center serves chtldren younger than 2 years of age, 0
                                                             otherwise
                                   URBAN                     Equals 1 if located in an urban area, 0 otherwise
                                   LOINCOME                  ProportIon of children from low-income families
                                   COST                      Total annual cost of center operation-rncludes    value of donated
                                                             labor, space, and supplies
                                   EDUC                      Years of formal education
                                   EXP                       Years of employment experience In early childhood education/
                                                             childhood development
                                   FULLTIME                  Equals 1 If considered working full time, 0 otherwise
                                   ADULTS                    Total number of adults employed by each center
                                   LEAVE                     Equals 1 if some or all teachers/atdes receive paid vacation leave, 0
                                                             otherwise
                                   REDFEE                    Equals 1 rf some or all teachers/aides eligible for reduced child care
                                                             fee, 0 otherwise
                                   HEALTH                    Equals 1 If some or all teacher/aides eligible for partly or fully paid
                                                             health insurance, 0 otherwise
                                   MIDWEST              __-- Equals 1 if in Midwest, 0 otherwise
                                   WEST                      Equals 1 if in West, 0 otherwise
                                   SOUTH                     Equals 1 If tn South, 0 otherwise




                                    Page 80                                          GAO/BRD-90-43BR     Early Childhood   Education
                                       Appendix N
                                       Technical Description  of GAO’s Economic
                                       Analysis of the Costs of HighQuality Early
                                       Childhood Education




Table IV.2: Multiple Regression Cost
Equation Estimates                     Dependent Variable= In (Total cost)
                                       Variable         Description
                                                            ~~                   Est coeff          Std error             T-stat     Mean value
                                       WAGET            Teacher avg.
                                                        wage/hr                        0.304             0 052              5.89               7 29
                                       WAGEA            Aide avg. wage/
                                                        hr                             0.127             0.062              2 05               5 11
                                       WAGED            Director avg
                                                        waaelmo                        0.165             0.038              4.36          2060 25
                                       RENT             Occupancy
                                                        cost/q foot                    0.055             0 007               7 40               3 05
                                       OCOST            Other cost/FTE
                                                        child                          0.362             0.024             15.01           1364 3.7
                                       CHILDREN         No. of FTE
                                                        children                       0.801             0.029             28.02               81 65
                                       CARAT10          z;;d-to-adult
                                                                                     -0.045              0.008             -5.41                4w
                                                                                                                                                 __
                                       GROUPSZ          Child group size               0.001             0 002               0.64              1678
                                       OUTPC            Outdoor
                                                        w;eCW/                         0 033             0.014               2.38               0 28
                                       NEW              New center
                                                        (less 5 yrs)                 -0.021              0 041             -0 50                0.12
                                       PROFIT           For-profitcenter             -0.030              0.040             -0.75                0 15
                                       HANDICAP         Handicapped
                                                        children                       0.017             0.174               0.10               0 06
                                        INFANTS         Serves Infants                 0.110             0.031               3.55               0 45
                                        URBAN            Located In an
                                                         MSA                           0 069              0.044              1 54               0 89
                                        LOINCOME         Low Income
                                                         Children                     -0.053              0 053            -1 01                0 28
                                        CONSTANT                                        4.719             0.271            17 398
                                                                                           ______
                                        COST             Total annual
                                                         center cost


                                        Number of Observations                              la7
                                        Adjusted R-Square                                 0 906
                                        F-Statistic                                      12096
                                        F-Statlstlc Slgnlflcance Level                     0 00

                                        Total cost IS the total annual cost of operating the center, lncludlng the value of donated services
                                        supplies, and space
                                        Total cost, the five Input price variables (WAGET, WAGEA, WAGED, RENT, and OCOST), and CHIL-
                                        DREN are all measured In logarithms The means reported for these variables In the last column repre-
                                        sent the mean values of these variables In levels.
                                        aThe mean CARAT10 IS lower than the average child-to-teacher ratio reported earlier for two reasons
                                        First, CARAT10 includes children of all ages, not just 4-year-olds Because the child-to-teacher ratlo
                                        typlcally IS higher for older children the inclusion of Infants and toddlers along with 4-year-olds tends to
                                        lower the overall center ratio Second, part-time workers are Included In the calculation of CARAT10




                                        Page 8 1                                                GAO/HRD904BR          Early Childhood    Education
                                                Appendix IV
                                                Technical Description  of GAO’s Economic
                                                Analysis of the Costa of High-Quality Early
                                                Childhood Education




Table IV.3 Estimates of Rearession Eauation for Teachers and Aides
Dependent Variable: Ln Waqe
                               Est. coefficients              Standard errors                       T-statistic                Mean value
Variable     Description      Teachers           Aides       Teachers        Aides              Teachers          Aides     Teachers      Aides
EDUC         Yrs of
             education            0 058          0.032           0 005           0.003               11.93         9.25          15.01               13.4c
EXP          Yrs of
             experience            0.033         0.022           0.003           0.004                9.53          6.03          5.77                3.08
EXP*EXP                          -0.001.       -0.001           0.0002          0.0002              -5.59         -3.01          58.33               20.44
FULLTIME     Work full
             time                 0.074          0 026           0 028
                                                                    .-           0.013
CARAT10      Child/
             adult ratto         -0 003          0.005           0.005           0.004              -0.64           1.51          4.80                 4.62
HANDICAP     Handicapped
                 -____             0 551 -       0.192           0.110           0.083                4.99          2.32          0.05                 0.05
ADULTS       No of
             adults                0 002       -0.001           0.0006          0.0004                2.91        -2.37          22.32               23.49
PROFIT       For-profit          -0.030        -0.075            0.023           0.021              -1.34         -3.64           0.17                016
LEAVE        Paid leave          -0.217        -0 122            0 045           0019               -4.83         -6.45           0.97                0.85
REDFEE       Reduced
             child fee           -0 111          0.014           0.017           0.013              -6.43           1.03           0.62                0.57
HEALTH       Pard
             health Ins.          0 109          0.141           0.020           0.015                5.31          9.25           0.80                0.68
L%lNCOME     Low
             Income              -0 223        -0.185            0.032           0.024              -7.03         -7.56            0.22                0.24
URBAN        Urban
             location              0 181         0.089           0.028           0.021                6.52          4.19           0.92                0.91
GDwEST                           -0 205        -0.090            0.024           0.018              -8.69         -5.12            0.35                0.35
WEST                             -0.147        -0.087            0.027           0.020              -5.41         -4.29            0.17      .         0.19
SOUTH                            -0 184        -0.051            0.025           0.019              -7.38         -2.65            0.34                0.29
CONSTANT                          0 994          1.088           0.096           0.058               10.33         18.91


                                                Number of Observattons                   12ao=       1423b
                                                Adjusted R-Square                    0351           0.225
                                                F-Statistic                           442           26.84
                                                F-Statlstlc Sigmflcance Level            ,000         000

                                                a Teachers
                                                b Aldes.




                                                 Page 82                                                GAO/HR.D-9043BR    Early Childhood       Education
Appendix V

Major Contributors to This Briefing Report


                   Fred E. Yohey, Jr., Assistant Director, (202) 245-9623
Human Resources    Deborah R. Eisenberg, Assignment Manager
Division,          Ellen Kehoe Schwartz, Evaluator-in-Charge
Washington, D.C.   Sandra Baxter, Evaluator
                   James Cosgrove, Economist
                   Luann Moy, Social Science Analyst
                   Joan Vogel, Evaluator (Computer Science)




(104636)           Page 03                               GAO/HUD-WBR    Early Childhood Education
Requests for copies O~GAO reports should be sent to:

US. General AccouMhg     Offke
Post office Box 6016
Gaithersburg, MkryIaM    20877

Telephone 2W27b62431.

The fmt five copies of each report are free. Additional   copies are
92.00 each.

There is 8 25% dtsamuf   on orders for 100 or more copies mailed to a
singleaddress.

Orders must be prepaM by cash or by check or money order made
out to the Superintendent of Doeuntents.
united   states
Generai Accounting   oifice
Washington,  D.C. 20648

officialBnsiness
Penalty for Private Use $%I@.