Child Care: How Do Military and Civilian Center Costs Compare?  (Letter
Report, 10/14/1999, GAO/HEHS-00-7).

Pursuant to a congressional request, GAO compared the cost of the
Department of Defense's (DOD) high-quality child development program
with the cost of comparable care in the civilian market, focusing on:
(1) identifying the objectives of the military child development program
and describing how it operates; (2) determining the full cost of
operating DOD's U.S. child development centers and the cost per
child-hour for center-based care; and (3) comparing the cost per child
in DOD's child development centers with the full cost of comparable
quality child care in the civilian market.

GAO noted that: (1) the primary objective of the military's child
development program is to help military families balance the competing
demands of family and military responsibilities by providing
high-quality child care at affordable rates; (2) DOD's child development
program is implemented in each service and provides several child care
options, including center care, family child care, and before- and
after-school programs; (3) to promote a high-quality child development
program, DOD requires that caregiver salaries meet certain prescribed
minimum levels and that caregivers across all military services complete
comprehensive child development training; (4) DOD is required by law to
maintain strict oversight of the health and safety standards of its
child development settings through inspections; (5) funding for the
program comes from parent fees as well as federal funds; (6) federal
funds go toward supplies, equipment, staff training costs, and some
staff salaries; (7) about $315 million in federal funds was obligated in
fiscal year (FY) 1998 for DOD's child development program; (8) DOD
allocated approximately 80 percent of this amount to child development
centers, 10 percent to family child care, and 10 percent to school-age
care; (9) GAO estimated that the cost to the Air Force of operating its
U.S. child development centers was approximately $81.4 million in FY
1997, and the estimated cost per child-hour in these centers was $3.86;
(10) labor costs--which include the salaries and benefits of child
development center caregivers, directors, and support staff--composed 75
percent of the estimated cost, with the majority of labor costs
representing the salaries and benefits of caregivers; (11) the estimated
cost per child-hour varied significantly for different age groups, from
$5.41 for infants, to $4.28 for toddlers, to $3.24 for preschoolers;
(12) child-hour costs are higher for younger children because quality
standards for young children require more caregivers per child; (13)
because almost half the children that the Air Force centers serve are
under the age of 3, the centers' total costs and costs per child-hour
are higher than if they served a smaller proportion of children under
the age of 3; (14) when adjusted for age distribution of the children,
the costs of high-quality care in Air Force and civilian centers were
not substantially different; and (15) the adjustment reduced the Air
Force cost per child-hour from $3.86 to $3.42, which is about 7 percent
higher than the cost in civilian centers.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  HEHS-00-7
     TITLE:  Child Care: How Do Military and Civilian Center Costs
	     Compare?
      DATE:  10/14/1999
   SUBJECT:  Day care centers
	     Comparative analysis
	     Child care programs
	     Cost analysis
	     Military dependents
	     Preschool education
	     Quality assurance
IDENTIFIER:  DOD Child Development Center Program
	     Air Force Child Development Program

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************

Cover
================================================================ COVER

Report to Congressional Requesters

October 1999

CHILD CARE - HOW DO MILITARY AND
CIVILIAN CENTER COSTS COMPARE? 

GAO/HEHS-00-7

Child Care Costs

(116016)

Abbreviations
=============================================================== ABBREV

  CIS - Caregiver Interaction Scale
  CPI - Consumer Price Index
  DOD - Department of Defense
  ECERS - Early Childhood Environment Rating Scale
  FTE - full-time equivalent
  GS - general schedule
  ITERS - Infant/Toddler Environment Rating Scale
  MCCA - Military Child Care Act
  NAEYC - National Association for the Education of Young Children
  OFP - Office of Family Policy
  TIS - Teacher Involvement Scale

Letter
=============================================================== LETTER

B-279806

October 14, 1999

The Honorable Trent Lott
Majority Leader
United States Senate

The Honorable Larry E.  Craig
United States Senate

The Honorable Paul D.  Coverdell
United States Senate

The demand for child care has increased dramatically in the past
several decades as the number of mothers who work outside the home
has grown.  About 39 percent of women with children under the age of
6 were in the labor force in 1975; by March 1997, that figure had
risen to 65 percent.  Recognizing the importance of child care
arrangements to all working families, the Congress streamlined
federal programs in 1996 and increased funding in fiscal year 1997. 
Proposals have been introduced in the 106\th Congress that address
the affordability and quality of care.  An important question
underlying these proposals is how much high-quality child care costs. 

Researchers and practitioners agree that high-quality child care
settings are those in which a sufficient number of well-trained
caregivers have positive interactions with children engaged in
developmentally appropriate activities in safe environments.  Both
the Congress and the administration have praised the high quality of
the child development program operated by the Department of Defense
(DOD) and identified it as a model for the rest of the nation. 
Therefore, to provide a benchmark cost estimate for the Congress as
it addresses these child care issues, you asked us to compare the
cost of DOD's high-quality child development program with the cost of
comparable care in the civilian market.  As agreed with your offices,
our objectives were to (1) identify the objectives of the military
child development program and describe how it operates, (2) determine
the full cost of operating DOD's U.S.  child development centers and
the cost per child-hour for center-based care, and (3) compare the
cost per child in DOD child development centers with the full cost of
comparable quality child care in the civilian market.\1

The military child development centers in our review included only
those of the Air Force because it was the only service whose centers
had all demonstrated high quality by meeting the accreditation
standards of the National Association for the Education of Young
Children (NAEYC).\2 Our cost estimate data for the Air Force centers
came primarily from that service and other government sources.  We
also surveyed all Air Force child development centers to collect cost
information that these government sources did not provide.  Our cost
estimates for civilian child care centers came from information
provided by a research study\3 of 150 child care centers identified
as providing high-quality care out of a sample of 401 centers in four
states.  Although not necessarily representative of all high-quality
child care centers, these data are the only source of cost
information available in an existing database.  Appendix I provides
additional details about our methodology and its limitations.  We
conducted our work between April 1998 and July 1999 in accordance
with generally accepted government auditing standards. 

--------------------
\1 The full cost estimate represents what it costs to operate a
center when all costs are counted.  This estimate includes both
expended costs (total cash costs for daily operations) as well as
other costs that the Air Force incurs but does not charge its
centers.  Expended costs include labor, staff education and training,
food, supplies, equipment, and utilities.  Air Force occupancy cost
(annual value of center space) is imputed and counted only as a
component of the full cost estimate.  Also included in the full cost
estimate are legal services and donations.  Donations include
services that volunteers provide the centers as well as donated goods
and cash.  Most of our full cost findings are expressed as the cost
per child-hour, a measure that standardizes the cost of care across
all types of centers, regardless of the number of children served or
the hours of operation.  Although we present our cost information in
dollars and cents, our findings are estimates and not as accurate as
this level of measurement implies.  We refer to our cost findings as
"complete" or "total" costs.  We discuss data quality issues and
other study limitations in app.  I. 

\2 NAEYC is the nation's largest association of early childhood
professionals.  Its purpose is to improve professional practice in
early childhood care and education and increase public understanding
of high-quality early childhood programs.  NAEYC administers, through
a voluntary system, an early childhood program accreditation process
designed to set standards of excellence in early childhood education. 

\3 Child care centers examined in Cost, Quality and Child Outcomes in
Child Care Centers, ed.  Suzanne W.  Helburn (Denver:  Economics
Department, University of Colorado at Denver, 1995) were selected
using a stratified random sample of 100 centers in each of four
participating states with approximately equal representation of
for-profit and nonprofit centers.  The four participating states were
California, Colorado, Connecticut, and North Carolina. 

   RESULTS IN BRIEF
------------------------------------------------------------ Letter :1

The primary objective of the military's child development program is
to help military families balance the competing demands of family and
military responsibilities by providing high-quality child care at
affordable rates.  DOD's child development program is implemented in
each service and provides several child care options, including
center care, family child care, and before- and after-school
programs.  To promote a high-quality child development program, DOD
requires that caregiver salaries meet certain prescribed minimum
levels and that caregivers across all military services complete
comprehensive child development training.  In addition, DOD is
required by law to maintain strict oversight of the safety standards
of its child development settings through inspections.  For example,
the military's child development facilities are subject to four
unannounced inspections a year.  Funding for the program comes from
parent fees as well as federal funds.  Center caregiver salaries are
paid primarily with parent fees.  Federal funds go toward supplies,
equipment, staff training costs, and some staff salaries.  About $315
million in federal funds was obligated in fiscal year 1998 for DOD's
child development program.  DOD allocated approximately 80 percent of
this amount to child development centers, 10 percent to family child
care, and 10 percent to school-age care. 

We estimated that the cost to the Air Force of operating its U.S. 
child development centers was approximately $81.4 million in fiscal
year 1997, and the estimated cost per child-hour in these centers was
$3.86.  Labor costs--which include the salaries and benefits of child
development center caregivers, directors, and support staff--composed
75 percent of the estimated cost, with the majority of labor costs
representing the salaries and benefits of caregivers.  The estimated
cost per child-hour varied significantly for different age groups,
from $5.41 for infants, to $4.28 for toddlers, to $3.24 for
preschoolers.  Child-hour costs are higher for younger children
because quality standards for young children require more caregivers
per child.  Because almost half the children that the Air Force
centers serve are under the age of 3, the centers' total costs and
costs per child-hour are higher than if they served a smaller
proportion of children under the age of 3. 

When adjusted for age distribution of the children, the costs of
high-quality care in Air Force and civilian centers were not
substantially different.  The adjustment reduced the Air Force cost
per child-hour from $3.86 to $3.42, which is about 7 percent higher
than the cost of care in civilian centers.  Another factor that
affects the overall cost per child-hour in Air Force centers is the
caregiver compensation rate.  Air Force centers pay their caregivers,
on average, about $1.04 more per hour than comparable civilian
centers do. 

   BACKGROUND
------------------------------------------------------------ Letter :2

      MILITARY AND CIVILIAN CHILD
      CARE SETTINGS
---------------------------------------------------------- Letter :2.1

Both military and civilian child care services are supplied by
providers operating in varied settings:  with center care, a child is
cared for in a nonresidential setting; with family child care, a
child is cared for in the home of a provider; and with in-home care,
a child is cared for in his or her own home.  Child care
centers--also known as child development centers in the military and
nursery schools or preschools in the civilian sector--are
nonresidential facilities.  Military centers are located on military
bases, while civilian centers are located in a variety of
establishments, including churches, schools, businesses, and public
agencies.  Family child care is provided to a small number of
unrelated children--typically fewer than six--in a provider's home. 
On military bases, the provider's home is military housing, and the
provider is a military spouse.  Civilian in-home care, such as that
provided by au pairs or nannies, usually is provided for the children
of the family that resides in the home.  Military families may make
arrangements independently for in-home care, but care by au pairs or
nannies is not a service offered by the military child development
program.  In addition, some military and civilian centers and family
child care homes also offer before- and after-school care. 

      CHILD CARE QUALITY AND COST
---------------------------------------------------------- Letter :2.2

Military and civilian child care providers are concerned about the
quality of the child care experience.  Researchers and professional
associations have addressed the concern for quality by identifying
two broad quality dimensions that pertain to all child care programs: 
children's daily interactions with their caregivers and structural
features of the child care environment.  Research has shown that
child-caregiver interactions are most closely linked with children's
development.  Thus, child care is considered high-quality when
caregivers are sensitive in responding to children's social behavior,
participate in their play and learning activities, and guide their
behavior in a positive manner.  Structural features include group
size, the number of children assigned to a team of caregivers in a
classroom; child-to-staff ratio, the number of children per caregiver
in a classroom; caregiver training; and the amount of floor space per
child.\4 Along with adequate space, resilient playground surfaces and
frequent staff and child hand washing are important health and safety
structural features of high-quality child care.  Another important
structural feature, particularly for infants and toddlers, is low
staff turnover, which researchers have tied to higher caregiver
compensation.  While structural features are not directly associated
with children's development, they support and facilitate positive
child-caregiver interactions.\5

Standards established by professional associations address both the
interactive and structural dimensions of child care quality.  For
example, both NAEYC's accreditation standards and the measures
included in the Early Childhood Environment Rating Scale--a widely
recognized tool used to assess the dimensions of quality in child
care centers--address the quality of child-caregiver interactions as
well as structural features of the child care environment.\6 A
civilian center's compliance with NAEYC and other professional
association standards is voluntary.  For the military child
development program, the military services elected to meet NAEYC
standards as a supplement to DOD's requirements for child development
and safety.\7

Thus, both DOD and NAEYC child care standards serve as a basis for
DOD's child development program. 

Using statistical measures for child-caregiver interactions and
structural features of centers, as well as for numerous other
variables, the Cost, Quality and Child Outcomes in Child Care Centers
study examined the relationship among the quality of care, the cost
of care, and the developmental progress of children in civilian
center-based child care.  Conducted by a team of economists and
developmental psychologists from four universities, the study had as
its fundamental objective explaining the quality of care prevalent in
child care centers by looking at the operation of child care markets. 
Among a broad range of findings, the study produced three that were
pertinent to our analysis: 

  -- High-quality center-based child care costs more, but not a lot
     more, than other center-based care.\8

  -- The quality of care in child care centers is related to several
     components of labor cost, including child-to-staff ratios,
     caregivers' wages, and caregivers' education and training. 

  -- Most center-based care, especially of infants and toddlers, is
     mediocre at best. 

The 401 centers in the study population were selected from lists of
licensed child care facilities in the four participating states. 
Although the states were not selected randomly, the study team
believed the child care centers in the four states to be
representative of early care and education programs in the United
States.  Centers were selected for data collection within each state
using a stratified random sample, with approximately equal
representation of for-profit and nonprofit centers.  Approximately
100 centers were selected in each state.  Of the 401 centers, 25 were
accredited by NAEYC.\9

--------------------
\4 The child-to-staff ratios for various group sizes that NAEYC
recommends for high-quality care are provided in app.  II. 

\5 See Child Care:  Use of Standards to Ensure High Quality Care
(GAO/HEHS-98-223R, July 31, 1998) for more information on structural
features of the child care environment.  State child care standards,
which are mandatory, primarily focus on the structural dimensions of
care.  The Maternal and Child Health Bureau, Public Health Service,
U.S.  Department of Health and Human Services, has developed
guidelines for child health and safety, but compliance is voluntary. 

\6 Child development researchers at the University of North Carolina
developed the Early Childhood Environment Rating Scale for child care
centers to use in self-assessment and for use in program evaluations
and other types of research.  This rating scale was used in the Cost,
Quality and Child Outcomes in Child Care Centers study.  See app.  I
for a more detailed discussion of the child care quality measures
used in our study. 

\7 Military child development centers are required to meet standards
of operation necessary for accreditation by an appropriate national
early childhood programs accrediting body. 

\8 The mean cost per child-hour for all 401 centers in the study
sample was $2.31 in 1992 dollars, using the same cost components
included in our cost estimate.  When the 1992 dollars are adjusted
for inflation to 1997 dollars, this cost estimate becomes $2.64. 
However, in addition to representing the total study sample of
centers of varying quality, rather than a subsample of high-quality
centers, this cost estimate differs from our estimate of the cost per
child-hour in civilian high-quality centers in two important ways. 
First, this estimate was based on enrollment, rather than the number
of children in attendance, as the measure of children served; second,
this estimate excluded three for-profit child care centers operated
at a place of business, which were considered outliers because they
tend to have high-quality care as well as high costs.  See app.  I
for additional details about our methodology. 

\9 See table I.1 for a comparison of NAEYC standards and the measures
used to select our subsample of high-quality centers. 

      QUALITY INITIATIVES IN THE
      MILITARY CHILD DEVELOPMENT
      PROGRAM
---------------------------------------------------------- Letter :2.3

To address reports of poor program quality and greater demand from
its military personnel for child care services, DOD took steps to
improve its child development program beginning in the 1980s.  In a
1982 report, we highlighted several problems in the program,
including unsafe child care facilities, a lack of DOD-wide program
standards, and inadequate training of the program's caregivers.\10 In
response to our report, DOD issued a Child Care Action Plan in 1983
that focused on improving DOD child development facilities by
replacing or renovating them.  It also addressed child-to-staff
ratios and group sizes and set minimum DOD-wide standards for the
operation of child development programs.  The plan became the basis
for DOD's child care policy for all its military services. 

In 1988, a series of congressional hearings on the military child
development program considered several allegations of child abuse at
military child development centers.  The hearings led to the
enactment of the Military Child Care Act (MCCA) of 1989, which
mandated further changes in the military's center program.\11 Most of
the provisions in the act were designed to improve the quality of
care, either by improving the quality of the caregivers themselves or
by ensuring that DOD standards are enforced.  While most of the
provisions in MCCA pertained to the military's child development
centers, DOD issued guidance in the 1990s extending key provisions,
such as inspections and training requirements, to family child care
and school-age programs.  In 1996, DOD took another step toward
implementing servicewide quality standards by directing the military
services to work toward accreditation by a national accrediting body
for all child development centers.\12 As of June 1999, 89 percent of
all military centers were accredited by NAEYC, while in the civilian
sector, where accreditation is voluntary, approximately 6 percent
were NAEYC-accredited. 

--------------------
\10 Military Child Care Programs:  Progress Made, More Needed
(GAO/FPCD-82-30, June 1, 1982). 

\11 The Military Child Care Act was revised and codified with the
enactment of P.L.  104-106, section 568(a)(1), February 10, 1996. 
The provisions are currently found in sections 1791-1798 of title 10,
USC. 

\12 In addition, a 1998 White House memorandum directed all federal
child care centers to be accredited by 2000. 

   DOD'S CHILD DEVELOPMENT PROGRAM
   IS STRUCTURED TO PROMOTE
   HIGH-QUALITY, AFFORDABLE CARE
------------------------------------------------------------ Letter :3

According to DOD, its child development program has several
objectives, including helping families balance military and family
responsibilities; improving the family's economic well-being; and
promoting children's cognitive, social, emotional, and physical
development in all DOD's child development program settings.  The
military's program includes a range of child care options to
accommodate the needs of working military personnel.  While the
Office of Family Policy (OFP) within the Office of the Secretary of
Defense is responsible for developing overall child care policy, each
military service issues its own regulations based on DOD policy and
is responsible for operating its own program.  OFP's child care
policies are designed to promote a qualified and stable workforce and
effectively enforce OFP's health and safety standards.  The
military's program is funded from a combination of federal
appropriations and parent fees. 

      DOD PROVIDES A RANGE OF
      CHILD CARE OPTIONS
---------------------------------------------------------- Letter :3.1

To meet the needs of working military parents, who often work
nonstandard hours, including shift work, nights, and weekends, DOD
offers several child care arrangements.  Specifically, the military
services offer three principal forms of child care:  child
development centers, family child care, and before- and after-school
programs ("school-age care").  Those eligible for the military's
child development program include active duty military personnel, DOD
civilian personnel, reservists on active duty or during inactive duty
personnel training, and DOD contractors.  First priority for DOD's
child development program is given to employed parents who are active
duty military and civilian personnel.  The military assists families
in finding at least one affordable child care option located either
on or off the military base. 

On-base centers were the original source of formalized child care on
military bases and still serve as the principal component of the
military's child development program.  DOD's center-based care serves
about 41 percent of all children participating in the military child
development program in about 800 centers worldwide.  DOD's child
development centers vary in size, with the largest serving about 300
children.  Most of the centers serve children aged 6 weeks to 5
years, with about 45 percent of the children under the age of 3. 
These centers typically operate from 6:00 a.m.  to 6:30 p.m., 5 days
a week.  Many centers also offer a part-day preschool program and
hourly care. 

DOD's family child care programs generally are provided by the
spouses of active duty military personnel living in government-owned
or -leased housing.  The military services serve about 35 percent of
the children in the military's program in approximately 10,000 family
child care homes in the United States and overseas.  Family child
care homes serve the same population as centers and provide services
for sick and special needs children.  In addition, they offer
extended-hour and weekend care to accommodate military shift work and
provide long-term care during military deployments.  According to DOD
officials, the flexible hours offered by family child care homes play
an important role in providing child care during nontraditional hours
and ensuring that parents are available for military duties on short
notice. 

All military services operate before- and after-school care programs
for children aged 6 to 12.  These programs, which offer care during
holidays and summer vacation, are housed in various facilities,
including child development centers, youth centers, and schools
located on military installations.  While in the past the military
services have focused more on center and family child care, the
services' child development program managers told us that they are
now looking to either expand the school-age program or improve its
quality.  According to one service manager, one service is trying to
do both. 

OFP develops overall child development policies that pertain to
centers, family child care, and school-age care.  Each military
service issues its own child care regulations that are based on DOD's
policies, publishes instructions that pertain to the specific
operations of its program, and administers its own program.  While
the services have the authority to issue child care regulations that
are more restrictive than OFP policies, most military services' child
development program managers told us that, generally, they do not
make policies more restrictive because doing so might increase the
cost of program operations.  Military base officials make decisions
about the operation and management of their particular child
development programs.  For example, officials determine whether to
offer center-based care, family child care, or both and whether to
expand existing programs or facilities.  Center directors are
responsible for the day-to-day operation of the centers, and a family
child care director at each base oversees the family child care
providers. 

      IMPROVED WAGES,
      COMPREHENSIVE TRAINING
      REQUIREMENTS, AND
      CENTRALIZED OVERSIGHT
      PROMOTE HIGH-QUALITY CARE
---------------------------------------------------------- Letter :3.2

DOD requires that caregiver salaries meet certain prescribed minimum
levels.  In addition, OFP policies follow statutory requirements,
which mandate specific training requirements and strict oversight of
DOD's child care standards to enhance the quality of its child
development program.  While most if these requirements relate only to
center-based care, DOD has comparable training requirements for
caregivers in its centers, family child care homes, and school-age
programs and carries out comparable inspections and similar
background checks on all its caregivers.  The services also offer
resource and referral programs that provide parents with information
on and referrals to all program components. 

         IMPROVED WAGES
-------------------------------------------------------- Letter :3.2.1

In 1990, to provide military centers with a more qualified and stable
workforce, DOD began paying center caregivers wages equivalent to
those of other employees on the same military base with comparable
training, seniority, and experience.  Before this policy change, most
military caregivers, like their civilian counterparts, were paid
minimum wage and received few benefits.  According to DOD officials,
improved wages for center caregivers reduced turnover significantly
and improved the quality of care by providing a more stable
workforce.  Staff turnover in the military program is now less than
40 percent annually, down from 300 percent on some military bases
during the 1980s.  According to OFP, current turnover is primarily
the result of the "normal transfer" of military personnel, whose
spouses work in child care, from one military base to another.\13
Research has shown that higher staff salaries are associated with
better-quality child development centers and that children who attend
centers with lower staff turnover develop better social and language
skills. 

--------------------
\13 During 1997, 27 percent of child care teachers and 39 percent of
assistant teachers left their jobs in civilian centers.  In that same
year, one-fifth of civilian centers reported losing 50 percent or
more of their teaching staff.  See M.  Whitebook and others, Worthy
Work, Unlivable Wages:  The National Child Care Staffing Study,
1988-1997 (Washington, D.C.:  Center for the Child Care Workforce,
1998), p.  8. 

         COMPREHENSIVE TRAINING
         REQUIREMENTS
-------------------------------------------------------- Letter :3.2.2

DOD's training program for its newly hired caregivers in centers,
family child care homes, and school-age programs requires that
caregivers first complete orientation training before they are
allowed to work directly with the children.  Center and family child
care providers must then complete 15 competency-based training
modules within 18 months of their start date in order to be retained. 
OFP used nationally recognized competency standards to develop these
modules, which focus on areas such as promoting a safe and healthy
learning environment, advancing physical and intellectual competency,
and supporting the social and emotional development of a child.\14
Caregivers for the school-age program must complete 36 hours of
training based on these competency modules within the first year of
work.  Finally, all center, family, and school-age caregivers must
attend 24 hours of training annually.\15 DOD's training requirements
apply uniformly across all military services. 

To improve the quality of training for caregivers, DOD is also
required by law to ensure that each child development center employs
a training and curriculum specialist, and DOD policy requires that
the specialist be a professionally qualified early childhood
educator.  According to the director of OFP, the training and
curriculum specialist provides training for center, family child
care, and school-age providers.  This person is responsible for
developing the program's curriculum, promoting developmentally
appropriate practices, instructing caregiver staff, and ensuring that
the staff demonstrate skills as a result of the required training. 

DOD has also initiated an "up or out" policy that links its training
program to wages and promotion.  DOD requires that center caregivers
complete all training modules to remain employed and to receive the
highest hourly wage available to DOD caregivers.  Because family
child care providers are independent contractors and generally set
their own fees, their wages are not tied to training requirements. 
However, family child care providers can lose their contract to
provide care if they do not complete the required and ongoing
training. 

--------------------
\14 The Council for Early Childhood Professional Recognition operates
the Child Development Associate National Credentialing Program. 
Focusing on the skills of early childhood and educational
professionals, the program is designed to provide performance-based
training and assessment of child care staff and family child care
providers.  The program represents a national effort to provide child
development credentials to qualified caregivers who work with
children from birth through age 5. 

\15 In the civilian sector, fewer than half of the states require any
training for newly hired center caregivers before they can begin
working, and only 11 require similar training for family child care
providers.  Forty-four states require some kind of annual ongoing
training for center caregivers, and 31 require annual ongoing
training for family child care providers. 

         CENTRALIZED OVERSIGHT
-------------------------------------------------------- Letter :3.2.3

DOD strictly enforces its child care standards, such as those
pertaining to safety issues.  The law requires DOD to ensure that
each center receives not fewer than four unannounced inspections a
year.  DOD's policy states that base personnel must conduct three of
the four unannounced inspections, including at least one
comprehensive health and sanitation inspection, one comprehensive
fire and safety inspection, and one inspection led by a
representative with authority to verify compliance with DOD child
care standards.  By law, a fourth inspection is carried out by
representatives from a higher level of command, including a child
development specialist.  This fourth inspection includes a review of
the center's curriculum, staff, and training; interviews with
parents; and an assessment of the safety and appropriateness of
indoor and outdoor equipment.  In addition to these quarterly
inspections, base personnel conduct monthly fire and food sanitation
inspections at centers.  If a center program is in compliance with
DOD standards, DOD issues a certificate to operate.  If the center is
not in compliance and the violation is life-threatening, the base
commander is required by law to take immediate action to correct the
problem.  In non-life-threatening cases, the violation must be
corrected within 90 days or the center is closed until the violation
is remedied. 

DOD policy also requires that family child care homes be inspected
quarterly.  These unannounced inspections include fire, safety,
health, and program inspections and ensure that child-to-staff ratios
are maintained.\16 According to DOD and military service officials,
these homes also are monitored monthly by a family child care
director who works at the base.  In addition, a multidisciplinary
team that includes a child development specialist randomly inspects a
representative sample of family child care homes.  Because most
family child care homes are located on military bases, they are not
subject to state licensing requirements.  However, as is the case
with centers, they may be closed if the provider fails to promptly
correct a violation. 

School-age programs, which are housed in child development centers,
youth centers, schools, or other base facilities, are subject to at
least one comprehensive and one unannounced inspection annually. 
These programs follow the same certification and remedy process for
violations that centers do. 

--------------------
\16 The size of the housing and age of the children determine the
maximum number of children a provider may care for.  Most providers
may care for no more than six children, including their own, under
the age of 8. 

      FEDERAL FUNDS HELP PAY FOR
      DOD'S PROGRAM AND REDUCE
      COST TO FAMILIES
---------------------------------------------------------- Letter :3.3

DOD's child development program is funded with a mix of federal funds
and parent fees.  Federal funds are provided directly to DOD's
center, family child care, and school-age programs in the form of
program subsidies.  These funds are used to reduce the cost of child
care for military families while improving the quality and
availability of care.  Increased demand for child care in the
military over the last 2 decades has influenced how the military
services spend their federal child care funding. 

Federal funds pay for supplies, equipment, training, and some
salaries for program staff, as well as for construction of new
centers and maintenance of existing centers.  These funds also pay
for some specific costs associated with family child care.  For
example, DOD pays the cost of administering the program, which
includes the salaries of the family child care directors.  In
addition, DOD provides indirect financial support through its
resource lending libraries, which loan equipment such as cribs, high
chairs, toys, books, and other supplies to these providers.  These
libraries, which also are paid for with federal funds, are located on
most military bases with child development programs.  A total of
approximately $315 million in federal funds was obligated in fiscal
year 1998 for military child development program operations,
including $253 million for center-based care, $31 million for family
child care, and $31 million for school-age programs (see table 1). 

                                Table 1
                
                    Funding Obligated for DOD Child
                 Development Program, Fiscal Year 1998

                         (Dollars in millions)

                                Family  School
                        Center   child    -age
Appropriation type           s    care    care          Total
----------------------  ------  ------  ------  ======================
Air Force
----------------------------------------------------------------------
O&M\a                    $78.8    $6.4   $13.1          $98.3
MilCon\b                  17.0      \b      \b           17.0

Army
----------------------------------------------------------------------
O&M                       68.6    10.1     8.4           87.1
MilCon                       0      \b      \b            0

Navy/Navy Reserve
----------------------------------------------------------------------
O&M                       62.6    11.5     7.3           81.4
MilCon                     5.0      \b      \b           5.0

US Marine Corps
----------------------------------------------------------------------
O&M                       13.8     2.5   2.6\c           18.9
MilCon                     7.0      \b      \b           7.0
======================================================================
Total                   $252.8   $30.5   $31.4          $314.7
----------------------------------------------------------------------
\a Operation and maintenance (O&M) funds pay for child development
program expenses such as the salaries and benefits of administrative
and management personnel, supplies, equipment, and utilities. 

\b Military construction (MilCon) funds are appropriated for
construction of new centers.  Family child care providers reside in
military housing, which is built with funds provided through
congressional appropriations for government housing.  School-age
programs not located in child development centers can be located in
youth centers, chapels, or dependent schools.  Funding for these
facilities is provided through different accounts. 

\c Includes funding for resource and referral services and oversight
requirements. 

Source:  DOD Budget Exhibit, Child Development, School-Age Care,
Family Centers, and Family Advocacy Program (PB-50), Jan.  1999. 

Parent fees for center and school-age care pay for most caregiver
wages in these two programs.  These fees are based on a sliding scale
that is determined according to family income in order to ensure that
military personnel with the lowest incomes can afford child care. 
For example, military families in the lowest two income categories
($0 to $23,000 and $23,001 to $34,000) paid $38 to $62 per child per
week for center care during the 1998-99 school year.  Families with
annual incomes over $55,000 paid $86 to $97 per child.  (See table
2.) In addition, to help ensure adequate funding for military child
care, federal law requires that DOD estimate total parent fees each
year and provide at least as much in federal funds to child
development centers.  Because family child care providers are
considered independent contractors by military bases, these providers
negotiate their fees directly with their customers. 

                          Table 2
          
               Parent Fees for Military Child
           Development Centers, School Year 1998-
                             99

                                    Range of
                                 weekly fees      Optional
                                  authorized     high-cost
Total family income              per child\a       range\b
------------------------------  ------------  ------------
$0-23,000                             $38-51        $43-54
23,001-34,000                          48-62         53-66
34,001-44,000                          59-74         65-79
44,001-55,000                          72-84         78-90
55,000+                                86-97        89-101
----------------------------------------------------------
\a Base commanders may authorize up to a 20-percent reduction of fees
charged for each additional child from the same family. 

\b An optional high-cost range may be used in areas where it is
necessary to pay higher wages to caregivers to compete in the local
labor market. 

Demand for child care has increased as a result of the increase in
the number of military personnel with families, women in the
military, and families with both parents in the military.  In
addition, the number of military spouses working outside the home
increased from 30 percent in 1970 to 62 percent in 1997.  In the
early 1990s, DOD established a formula for estimating the need of its
military families for child care services that was based on the
number of children up to age 12 in military families whose parents
worked outside the home and needed some type of child care. 
According to DOD, as of 1998, the military services were meeting
about 56 percent of the projected need for child care.  DOD's goal is
to meet 80 percent by 2005.  The director of OFP stated that DOD
believes that the remaining 20 percent of military families with
young children will not request child care either because the parents
have alternating work schedules or because relatives care for their
children. 

During fiscal years 1985 through 1998, the military services built
about 208 new centers to accommodate the child care needs of their
military families.  Military personnel generally show a preference
for placing their children in centers, in part because center fees
are generally lower than those charged by family child care
providers.  OFP's director told us that because of the high cost of
constructing and operating child development centers, the high cost
of caring for young children in these centers, and military parents'
need for child care, most military services are trying to encourage
the use of more family child care providers.  For example, in the
1990s, the Army, Navy, and Marine Corps began offering subsidies in
the form of direct cash payments to family child care providers so
that they could charge fees comparable to center fees.\17

While the Marine Corps primarily target providers caring for infants
and toddlers for subsidies, the Army and Navy offer subsidies to
providers caring for children up to 5 years of age.  The Army targets
some of its subsidies to providers that are caring for special needs
children and to those that are open extended hours.  The Navy program
manager told us that about half of Navy base commanders are offering
family child care providers some amount of subsidy. 

--------------------
\17 According to the chief of the Air Force Family Member Program,
the Air Force's primary child care goal is to ensure that all parents
that need child care have a child care option available to them.  She
said the Air Force does not have sufficient resources to increase the
amount of care available and to reduce the cost of care.  She
believes that as more Air Force centers are built and, consequently,
more center care is available, family child care providers will be
forced to bring their fees more in line with center fees. 

   THE HIGH PROPORTION OF YOUNG
   CHILDREN IN AIR FORCE CENTERS
   RESULTS IN HIGHER COSTS
------------------------------------------------------------ Letter :4

Air Force child development centers are similar to centers in the
other military services.  They serve children in various age groups
and are required to follow servicewide regulations designed to
promote high-quality care.  The estimated cost to the Air Force of
operating its U.S.  center program was about $81.4 million in fiscal
year 1997, and the estimated cost per child-hour in Air Force centers
was $3.86.  Labor costs constituted three-fourths of this cost and
differed depending on the age of the child, with significantly higher
labor costs per child-hour for young children.  In addition, a large
proportion of the children at Air Force centers are in the youngest
age groups, increasing the centers' total child care costs and cost
per child-hour. 

      CENTERS' LABOR COSTS
      CONSTITUTE MAJORITY OF AIR
      FORCE ESTIMATED COST
---------------------------------------------------------- Letter :4.1

We estimated that the Air Force's total cost for operating its child
development centers in the United States was approximately $81.4
million in fiscal year 1997, the most recent year for which data were
available.  Labor costs accounted for 75 percent of the total
estimated cost.  The cost of labor in Air Force centers includes
costs for direct labor (caregivers) and indirect labor (all other
staff).  Salaries and benefits of caregivers accounted for about 52
percent of the total cost and 70 percent of labor costs.  The Air
Force employs about 1,800 caregivers in its U.S.  centers and about
600 other employees, including directors, administrative staff, and
cooks.  Many of the centers have an assistant director as well as a
director, and 19 percent also have annex directors that serve
children in additional space outside the main center building.  In
addition, almost all centers have a training and curriculum
specialist who provides services for center caregivers and management
staff, and most centers have their own cooks who prepare breakfast
and lunch for the children.\18

About 64 percent of Air Force centers (93) are located in the United
States, with the remaining located on military bases overseas.\19
U.S.  centers are housed on military bases, in either newly
constructed facilities built specifically for use as child
development centers or buildings that were renovated to conform to
DOD child care standards.  We estimated that the cost of providing
this space constitutes the second largest component of center
cost--about 10 percent of the total cost.  Supplies used by the
center, such as classroom and administrative materials, were the
third largest component, accounting for about 7 percent.  (See table
3.)

                          Table 3
          
           Estimated Full Cost of Air Force Child
           Development Centers, Fiscal Year 1997

                                                Percentage
                                                  of total
Cost component                    Total cost          cost
------------------------------  ------------  ------------
Labor (wages and benefits)
----------------------------------------------------------
Direct labor                     $42,655,050         52.40
Indirect labor                    18,763,938         23.05
Supplies                           5,710,243          7.01
Food                               3,939,678          4.84
Utilities                          1,537,571          1.89
Equipment                            582,165          0.72
Total expended costs             $73,188,645         89.91
Occupancy                          7,936,879          9.75
Donations                            237,281          0.29
Legal services                        46,295          0.06
Noncash costs                     $8,220,455         10.10
==========================================================
Full cost                        $81,409,100         100\a
----------------------------------------------------------
\a Percentages do not equal 100 because of rounding. 

Source:  GAO analysis of Air Force center data. 

--------------------
\18 Training and curriculum specialists in the Air Force also serve
family child care providers and caregivers in the school-age program. 

\19 The number of Air Force child development centers reported is
based on the response to our survey. 

      CARING FOR YOUNGER CHILDREN
      INCREASES AIR FORCE LABOR
      COSTS PER CHILD-HOUR
---------------------------------------------------------- Letter :4.2

The overall estimated cost per child-hour in Air Force centers was
$3.86 in fiscal year 1997.\20 However, the cost varied for different
age groups, with younger children costing more.  For example, our
estimated cost for children aged 6 to 12 months was about $5.40 per
child-hour, compared with about $3.96 for toddlers aged 24 to 36
months and $3.23 for full-day preschoolers aged 3 to 5 years.  (See
fig.  1.)

   Figure 1:  Estimated Cost per
   Child-Hour at Air Force
   Centers, by Age Group, Fiscal
   Year 1997

   (See figure in printed
   edition.)

Source:  GAO analysis of Air Force center data. 

The higher cost per child-hour is associated with the younger age
groups because the child-to-staff ratios are lower for younger
children, which results in fewer children per caregiver.\21 Because
centers have to hire more caregivers for younger children, the
overall cost of labor per child-hour increases.  Our analysis shows
that in U.S.  Air Force centers there are about three infants per
caregiver, about four and one-half toddlers per caregiver, and about
eight preschoolers per caregiver.  Consequently, direct labor costs
are higher per child-hour for the youngest children.  (See app. 
III.) While more staff are assigned to the groups of younger
children, the average caregiver compensation rate, about $11.20 per
hour (including salaries and benefits), did not differ substantially
across the various age groups. 

Just as the other military services do, the Air Force employs a high
percentage of young people with young children.\22 Because many of
these families prefer center-based care, a high proportion of
children served in Air Force centers are young children.  Nearly half
are under 3 years of age:  38 percent are toddlers and 10 percent are
infants.  Moreover, the cost of toddler care and the cost of
preschooler care represent about the same proportion of the total
cost of child care--about 42 percent--even though preschoolers
constitute half the children.  In addition, even though infants
account for only 10 percent of all the children in Air Force centers,
the cost of their care accounts for about 14 percent of the overall
cost of child care.  (See table 4.) Thus, the centers' total cost and
their cost per child-hour are higher than if they served a smaller
proportion of children under the age of 3. 

                          Table 4
          
          Age Group Distribution and Cost of Care
           at Air Force Centers, Fiscal Year 1997

                                                Percentage
                                  Percentage      of total
Children's age                   of children          cost
------------------------------  ------------  ------------
Infants
----------------------------------------------------------
6 weeks-6 months                        3.38          4.76
6-12 months                             6.80          9.52
==========================================================
Subtotal                               10.18         14.28

Toddlers
----------------------------------------------------------
12-24 months                           16.20         19.81
24-36 months                           22.05         22.67
==========================================================
Subtotal                               38.25         42.48

Preschoolers
----------------------------------------------------------
3-5 years                              46.62         39.07

Part-day preschoolers
----------------------------------------------------------
3-5 years                               3.77          3.26
==========================================================
Subtotal                               50.39         42.33

School-aged children
----------------------------------------------------------
6 years and over                        1.19           .90
==========================================================
Total                                    100           100
----------------------------------------------------------
Note:  Percentages do not equal 100 because of rounding. 

Source:  GAO analysis of Air Force center data. 

--------------------
\20 The annual cost per child in an Air Force child development
center in the United States was about $8,028 in fiscal year 1997. 

\21 DOD's standards for child-to-staff ratios generally follow the
NAEYC standards (see app.  II).  The child-to-staff ratios that we
estimated are based on actual attendance reported by the centers in
fiscal year 1997.  According to the Air Force Child and Youth
Services manager, the observed ratios exceed the DOD standards (fewer
children per staff) because of the child absentee rate in the
centers.  She noted that, on average, about 12 percent of children
are absent each day from Air Force centers because of Air Force
personnel work schedules, illness, and family vacations.  For
example, many Air Force personnel often fly an average of 3 days a
week and are home the rest of the week.  In addition, military
personnel receive regular "home leave" and may keep their children
home during these days. 

\22 Thirty-three percent of Air Force personnel are under the age of
26.  In addition, 38 percent of employed Air Force spouses have
children under the age of 2. 

   COST PER CHILD-HOUR IS SIMILAR
   IN AIR FORCE AND CIVILIAN
   CENTERS WHEN ADJUSTED FOR AGE
   DISTRIBUTION
------------------------------------------------------------ Letter :5

The hourly cost for care is about 20 percent higher in Air Force
centers than in civilian centers of comparable quality because the
Air Force's labor costs are higher.  To a great extent, differences
in the age distribution of the children served explained the
difference in the centers' labor costs:  a higher proportion of Air
Force center children than of civilian center children are under the
age of 3.  When we adjusted the Air Force estimated cost to account
for the difference in the children's age, the overall Air Force cost
per child-hour decreased $3.86 to $3.42, which is about 7 percent
higher than civilian centers' cost per child-hour.  Air Force centers
also compensate their caregivers more per hour than civilian centers. 

The civilian centers' overall cost per child-hour was $3.19.\23 Table
5 breaks down the per-child-hour cost components--overall and for
each age group--showing that the most costly component in both Air
Force and civilian centers' overall cost per child-hour was the cost
of labor.  Labor represented 75 percent of the cost per child-hour in
Air Force centers and 67 percent in civilian centers.  The greatest
difference between the individual cost components of the Air Force
and civilian centers was also in the cost of labor--a difference of
78 cents per child-hour.  The other cost components differed by less
than half that amount.\24

                                         Table 5
                         
                            Unadjusted Components of Costs per
                          Child-Hour, by Age Group, Fiscal Year
                                           1997

                         Air Force                                Civilian
           --------------------------------------  --------------------------------------
                                   School                                  School
                                    -aged                                   -aged
Cost       Infa  Toddle  Preschoo  childr  Overal  Infa  Toddle  Preschoo  childr  Overal
component   nts      rs      lers      en       l   nts      rs      lers      en       l
---------  ----  ------  --------  ------  ------  ----  ------  --------  ------  ------
Labor
-----------------------------------------------------------------------------------------
Salaries,  $4.6   $3.32     $2.27   $2.04   $2.91  $3.9   $3.33     $1.88   $1.74   $2.12
 wages,       3                                       5
 and
 benefits
Training     \a      \a        \a      \a      \a  0.02    0.02      0.01    0.01    0.01
 costs
=========================================================================================
Subtotal   $4.6   $3.32     $2.27   $2.04   $2.91  $3.9   $3.35     $1.89   $1.75   $2.13
              3                                       7
Occupancy  0.38    0.37      0.38    0.42    0.38  0.46    0.46      0.36    0.38    0.38
 \b
Food          0    0.21      0.21    0.18    0.19     0    0.13      0.12    0.09    0.12
Other\c    0.39    0.37      0.37    0.29    0.37  0.35    0.32      0.28    0.26    0.28
=========================================================================================
Total      $5.4   $4.27     $3.23   $2.93   $3.85  $4.7   $4.26     $2.65   $2.48   $2.91
 expended     0                                       8
 costs
Donations  0.01    0.01      0.01    0.01    0.01  0.37    0.26      0.30    0.18    0.28
 \d
=========================================================================================
Total      $5.4   $4.28     $3.24   $2.94   $3.86  $5.1   $4.52     $2.95   $2.66   $3.19
 cost         1                                       5
-----------------------------------------------------------------------------------------
\a Air Force training costs include compensation of the training and
curriculum specialist employed by each child development center. 
This cost was not broken out separately but is included in the Air
Force cost of labor. 

\b This is the estimated cost of space.  For Air Force child
development centers, occupancy cost is imputed because centers are
not actually charged for the cost of space.  The cost of construction
is assumed by Air Force bases and paid out of military construction
funds.  For civilian centers, occupancy cost may include the cost of
utilities and repair and maintenance services. 

\c Includes cost of utilities, legal services, liability insurance,
taxes, supplies, equipment, and payments to a parent company, if
incurred. 

\d Donations include services that volunteers provide the child
development centers, as well as donated goods and cash. 

Source:  GAO analysis of Air Force and civilian center data.  The
fiscal year 1992 civilian center data were escalated to fiscal year
1997 using the Consumer Price Index. 

The difference in Air Force and civilian centers' labor costs is
explained in large part by the difference in the age distributions of
the children they serve.  Most of a child care center's labor cost is
determined by the number of caregivers employed.  When centers follow
professional standards that require fewer children per caregiver for
young children, centers employ more staff to provide care for younger
children than for older children.  Thus, the higher the proportion of
young children in the centers' care, the higher their overall labor
costs per child-hour.  As noted, the Air Force centers served a high
proportion of young children.  In fact, as table 6 shows, a higher
percentage of Air Force center children (48 percent) than of civilian
center children (15 percent) were under 3 years of age.  Because Air
Force centers served a higher proportion of young children than
civilian centers did, their overall labor costs per child-hour were
higher. 

                                Table 6
                
                Centers' Proportion of Children, by Age
                        Group, Fiscal Year 1997

                          (Numbers in percent)

                                                           Air  Civili
Children's age\a                                         Force      an
------------------------------------------------------  ------  ------
Infants
----------------------------------------------------------------------
6 weeks-6 months                                          3.38      \b
6-12 months                                               6.80      \b
======================================================================
Subtotal                                                 10.18    4.99

Toddlers
----------------------------------------------------------------------
12-24 months                                             16.20      \b
24-36 months                                             22.05      \b
======================================================================
Subtotal                                                 38.25   10.39

Preschoolers
----------------------------------------------------------------------
3-5 years                                                46.62   72.95

Part-day preschoolers
----------------------------------------------------------------------
3-5 years                                                 3.77      \b

School-aged children
----------------------------------------------------------------------
6 years and over                                          1.19   11.68
======================================================================
Total                                                      100     100
----------------------------------------------------------------------
Note:  Percentages do not equal 100 because of rounding. 

\a Categories for this column apply only to Air Force center data. 
See note in app.  I concerning age group assignment and measurement
of children's ages in the civilian center data. 

\b Data for civilian centers were not broken down into this
subcategory. 

Source:  GAO analysis of Air Force and civilian center data. 

In view of the importance of children's age as a factor in the
overall cost per child-hour, we adjusted the Air Force cost estimate
to take account of the age distribution differences between the Air
Force and civilian centers' child populations.  To do so, we
calculated what the Air Force centers' costs would have been had the
age distribution of the children they served been the same as that of
the civilian centers.  The result was that the difference in the
centers' overall cost of labor per child-hour was reduced from 78
cents to 35 cents.  Much of the 35-cent difference in the cost of
labor that is left can be attributed to the higher compensation that
Air Force centers pay caregivers--an average hourly compensation rate
of $11.20 (salaries and benefits), which is $1.04 higher than the
rate paid at civilian centers. 

Finally, substituting the age distribution of the civilian centers'
children for that of the Air Force centers' children reduced the
overall cost per child-hour from $3.86 to $3.42, a decline from about
20 percent to about 7 percent.  Thus, when we adjusted for age
distribution, the cost of high-quality child care in Air Force and
civilian centers was not substantially different. 

--------------------
\23 The annual cost per child for the civilian child care centers was
about $6,635 in fiscal year 1997. 

\24 The 27-cent difference in the dollar value of donations the
centers received was the second greatest component cost difference. 
Air Force centers received comparatively fewer donations,
representing less than 1 percent of their overall cost per
child-hour.  However, donations constituted 9 percent of civilian
centers' cost per child-hour and consisted mostly of center space and
volunteers' services. 

   AGENCY COMMENTS
------------------------------------------------------------ Letter :6

DOD commented on a draft of this report and generally concurred with
our findings.  However, DOD raised three issues about our research
methods and how they might have affected the difference we found in
Air Force and civilian centers' overall cost per child-hour.  While
the issues DOD raised are important study design considerations, the
magnitude and direction of any cost differences that would result
cannot be estimated.  Moreover, as we concluded, the cost difference
between Air Force and civilian centers is not substantial, and any
shift in cost differences resulting from the issues DOD raised is
unlikely to change that conclusion. 

First, DOD considered the measure we used for the number of children
served by Air Force centers dissimilar from the one we used for
civilian centers and suggested that the Air Force center cost per
child-hour was somewhat high in comparison to the civilian center
cost because of that dissimilarity.  However, we used the same
measure--attendance--for the number of children served in both Air
Force and civilian centers.  For Air Force centers, attendance was
based on counts conducted every hour of every day at each center; for
each civilian center, attendance was based on a 1-year extrapolation
of attendance reported for the day on which the Cost, Quality and
Child Outcomes in Child Care Centers study team visited to collect
data.  Presumably, the Air Force center attendance data were more
precise than the extrapolations we used for the civilian centers. 
But we have no basis for determining whether more precise civilian
center attendance data, if they had been available, would have been
higher or lower than the extrapolated data.  Thus, there is no basis
for concluding whether more precise civilian center data would have
decreased or increased the difference between estimated Air Force
center costs and estimated civilian center costs. 

Second, DOD argued that the Air Force and high-quality civilian
centers were not comparable in terms of quality because of the method
we used to select the civilian centers.  DOD suggested that a
difference in the centers' quality of care resulting from the sample
selection procedure might explain the cost difference that we found. 
We selected civilian centers for our study from the Cost, Quality and
Child Outcomes in Child Care Centers study sample on the basis of a
score of 4.5 or higher on either an overall index of quality or a
second index.\25 DOD pointed out correctly that a score of 4.5 to 5
on an overall index of quality was considered mediocre in the Cost,
Quality and Child Outcomes in Child Care Centers study of April 1995. 
However, the goal in identifying civilian centers for our comparative
cost analysis was to select civilian centers of a quality comparable
to the Air Force centers.  To achieve that goal, we had to find a way
of measuring child care quality in both sets of centers.  The only
readily available information on the quality of care in the Air Force
centers was the fact that all of those centers had been accredited by
NAEYC.  However, data on the quality of care in all 401 civilian
centers in the Cost, Quality and Child Outcomes in Child Care Centers
study sample had been collected using the two indexes noted above--an
overall index of quality and a second index.  A small number of these
civilian centers (25) also were accredited by NAEYC.  We looked at
the NAEYC-accredited civilian centers' scores on the two indexes of
quality as surrogate measures for the Air Force centers' scores on
these indexes.  We found that only 11 of the 25 had scored 5 or
higher, while 4 had scored below 4.5.  Rather than use the lowest
score as our cutoff point for identifying high-quality centers, we
selected the more conservative score of 4.5 or higher as the range in
which NAEYC-accredited centers' scores likely would fall on the two
indexes of child care quality.  Thus, we were satisfied that using a
score of 4.5 or higher on either of the two indexes to select
high-quality civilian centers for our analysis provided a subsample
of centers comparable in quality to the Air Force centers. 

Finally, DOD asserted that the cost per child-hour for the entire
military child development system--child development centers, family
child care, and school-age care--might have been less than the cost
per child-hour for the Air Force child development center program. 
Unfortunately, the detailed cost data for all military services and
all types of care that we needed in order to estimate the cost per
child for the entire system were not available for either the
military or the civilian child development systems. 

DOD also had several technical comments, which we incorporated in the
report where appropriate.  DOD's comments are included in appendix
IV. 

--------------------
\25 See app.  I for a discussion of our sample selection procedures. 

---------------------------------------------------------- Letter :6.1

As agreed with your offices, unless you publicly release its contents
earlier, we will make no further distribution of this report until 30
days after its issue date.  At that time, we will send copies to the
Honorable William S.  Cohen, Secretary of Defense, and congressional
committees with an interest in this matter.  We also will make copies
available to others on request. 

If you or your staff have any questions about this report, please
contact me on (202) 512-7215 or Karen Whiten, Assistant Director, on
(202) 512-7291.  Other GAO contacts and staff acknowledgments are
listed in appendix V. 

Cynthia M.  Fagnoni
Director, Education, Workforce, and
 Income Security Issues

SCOPE AND METHODOLOGY
=========================================================== Appendix I

This appendix discusses in more detail the scope and methodology for
determining the cost per child of Air Force and civilian center-based
child care. 

   STUDY SCOPE
--------------------------------------------------------- Appendix I:1

Our cost analysis did not cover the entire military child development
program because we could not find detailed cost data for all services
or for all types of care.  The military child development program is
implemented throughout the Department of Defense (DOD), but the
military services conduct program cost accounting and reporting
independently for the child development programs they operate.  Each
of the military services' child development program managers has some
discretion over the cost data collected and maintained by the
program.  So, when we searched for sources of program cost data at
DOD, we found only two reports about child development program
funding that all four services submitted in comparable format:  (1)
the Appropriated and Nonappropriated Fund Expense Summary for 1997
(the "7000.12 report") and (2) DOD's Program Budget Exhibit (known as
the PB-50) for its child development program.  However, neither of
these reports contained information at the level of detail needed to
estimate and analyze the full cost of child development program
operations in all four military services, particularly for those cost
components we expected to be the most important:  the cost of labor
and space.  Moreover, none of the services collected information on
the salaries and benefits of family child care providers. 

To design the comparative cost analysis, we also had to identify
sources of total cost information for civilian child care.  After an
extensive literature review, we found that only the Cost, Quality and
Child Outcomes in Child Care Centers study collected complete cost
information for civilian child care programs, and that this study was
limited to child care centers.  Although a study of civilian family
child care was under way, that study's data file was not released in
time to use it in our study.  Thus, we narrowed the study scope to
child care centers, the child care type for which the most complete
cost data for both military and civilian child care were available. 
Because the study focus was the cost of high-quality child care, we
limited the scope of the military cost analysis to the Air Force
child development center program, the only one of the four military
programs in which all centers had achieved National Association for
the Education of Young Children (NAEYC) accreditation.  We treated
NAEYC accreditation as the measure of quality attained by the Air
Force centers and selected a subsample of high-quality civilian
centers from the Cost, Quality and Child Outcomes in Child Care
Centers study population for our comparative cost analysis, using an
overall index of quality for infant, toddler, and preschool learning
environments that the Cost, Quality and Child Outcomes study team had
developed for their work. 

We did not attempt to independently assess the quality of the DOD or
civilian centers included in our study.  However, to determine
whether the Air Force child development centers were comparable to
the subsample of high-quality civilian centers, we examined the
similarity of the NAEYC accreditation standards and the Cost, Quality
and Child Outcomes study team's overall index of quality.  To compare
the accreditation standards and the index of quality, we adapted 15
indicators of child care quality specified by the National Research
Council's Panel on Child Care Policy.\26 The Panel developed these
indicators to compare the provisions of six sets of professional
standards on child care quality, including the NAEYC standards and
the Early Childhood Environment Rating Scale (ECERS) used in this
study. 

While not identical, the standards NAEYC uses to assess centers for
accreditation and the dimensions of quality measured by the overall
index are analogous.  Both NAEYC standards and the overall quality
index measure child-caregiver interactions and structural features of
the child care environment.  NAEYC's National Academy of Early
Childhood Programs developed the accreditation standards by reviewing
child development research, child development program standards in
localities, and the judgments of 175 specialists in the early
childhood field.  The accreditation standards address the quality of
interactions between children and caregivers as well as structural
features such as child-to-staff ratios and group size. 

The Cost, Quality and Child Outcomes study team's overall index of
quality integrates measures from four scales for early childhood
learning environments:  (1) ECERS, (2) the Infant/Toddler Environment
Rating Scale (ITERS), (3) the Caregiver Interaction Scale (CIS), and
(4) the Teacher Involvement Scale (TIS).  ECERS is designed to assess
the surroundings created for children and adults in an early
childhood setting.  It is used in one classroom at a time for groups
of children 2 to 6 years old.  Ratings are based on the situation
observed in the classroom or reported by the teacher, rather than on
plans for the future.  ITERS is an adaptation of ECERS by its authors
especially for use with infants and toddlers.  The CIS measures
teacher involvement and teacher style, producing scores for
"sensitivity," "harshness," "detachment," and "permissiveness." The
TIS also measures child-caregiver interactions, focusing on how a
caregiver interacts with children during periods of observation.  The
caregiver's degree of involvement is scored on a 6-point scale,
ranging from "ignore" to "simple" to "intense."

The NAEYC standards incorporate all 15 of the indicators of quality
developed by the Panel on Child Care Quality.  ECERS and ITERS
together account for 11 of the indicators among their scales and 2
more, "staff/child ratio" and "group size," could be calculated from
the study team's data file.  The indicator "potential for forming an
affectionate relationship with a familiar caregiver," which ECERS and
ITERS do not measure directly, is captured by the CIS and the TIS. 
One indicator, which pertained to a requirement for caregiver
training, was not measured by the scales in the Cost, Quality and
Child Outcomes study team's overall index of quality or by other data
collection instruments used in their study.  However, because all but
1 of the 15 quality indicators are represented in the accreditation
standards and index of quality, we considered the definition of child
care quality used in both assessment tools consistent and the quality
of care in the two sets of centers comparable.  See table I.1 for the
results of the comparison. 

                                        Table I.1
                         
                          Comparison of NAEYC Standards and the
                          Overall Index of Quality Developed for
                           the Cost, Quality and Child Outcomes
                                          Study

                                               Scales in the overall index of quality
                                           ----------------------------------------------
                                    NAEYC
Indicators\a                    standards       ECERS       ITERS         CIS         TIS
-----------------------------  ----------  ----------  ----------  ----------  ----------
Fosters potential for forming           x                       x           x           x
 an affectionate relationship
 with a familiar caregiver.
Encourages frequent positive            x           x           x           x           x
 interaction between
 caregiver and children.
 Caregivers are responsive,
 positive, accepting, and
 comforting.
Requires caregiver training             x
 related to child
 development.
Provides opportunities for              x           x           x
 caregiver training.
Specifies a maximum group               x          \b          \b          \b          \b
 size.
Specifies a child-to-staff              x          \b          \b          \b          \b
 ratio.
Provides curriculum that                x           x           x
 encompasses both
 socioemotional and cognitive
 development.
Provides children                       x           x           x
 opportunities to select
 activities.
Values experience with                  x                       x
 cooperative group process.
Provides structured but not             x           x           x
 overly rigid curriculum.
Encourages recognition and              x           x           x
 appreciation of children's
 culture.
Provides child-oriented                 x           x           x
 physical environment.
Provides orderly and                    x           x           x
 differentiated physical
 setting.
Provides for parental                   x           x           x
 involvement.
Encourages parent/staff                 x           x           x
 conferences and
 communication.
-----------------------------------------------------------------------------------------
\a Adapted from indicators developed by the National Research
Council's Panel on Child Care Quality. 

\b Measured by other Cost, Quality and Child Outcomes study team data
collection instruments. 

We also examined the scores that the 25 accredited centers in the
Cost, Quality and Child Outcomes study population received on the
overall index of quality and a second index that combined scores on
the ECERS and ITERS indexes.  Of the 25 accredited centers in the
entire study sample, 21 met the criteria for inclusion in the
high-quality subsample and were included among the 150 selected.\27
That is, 21 accredited centers had scores on the overall index of
quality or the ECER/ITER index of 4.5 or higher.  Four accredited
centers whose scores were less than 4.5 on both indexes were not
selected for the high-quality subsample, suggesting that the measures
of child care center quality employed by the overall index of quality
and the ECER/ITER index are slightly more stringent than NAEYC
standards. 

--------------------
\26 See National Research Council, Panel on Child Care Quality, Who
Cares for America's Children?  (Washington, D.C:  National Research
Council, 1990), app.  B, "Professional Standards for Early Childhood
Programs," pp.  324-39. 

\27 For purposes of our analysis, we considered centers that scored
4.5 or more on the 7-point index to be high-quality. 

   LIMITATIONS OF THE COST
   ANALYSIS
--------------------------------------------------------- Appendix I:2

Because there was no single data set for our analysis, we used
multiple sources of existing data and a questionnaire we developed. 
The fact that most data had been collected for purposes other than
our study resulted in some limitations.  Therefore, although we
present our cost information in dollars and cents, our findings are
estimates and are not as accurate as this level of measurement
implies.  The following measurement and analysis issues should be
considered in reviewing our findings. 

      DOD AND AIR FORCE COST DATA
      REQUIRED SUPPLEMENTATION
------------------------------------------------------- Appendix I:2.1

The Air Force Semiannual Child Development Center Report and the
Appropriated and Nonappropriated Fund Expense Summary for 1997 are
administrative cost reports developed for routine Air Force planning
and budgeting.  Therefore, several of the data items of interest for
our study had to be supplemented with data from other sources or
imputed.  For example, the semiannual report provided the salary
grade of child development center civil service employees paid under
the Civil Service general schedule (GS) but not the step or
locality.\28 The expense summary reported the cost of supplies
purchased by the child development centers, but the cost item
included bulk supplies for the family child care and school-age
programs as well.  One-half of the Air Force installations did not
report food costs on a more detailed expense summary provided by the
Services Directorate of the Air Force Family Member Program.  For
these and other items, we supplemented the data provided by DOD with
data from our survey or additional sources.  The Air Force data used
in our cost estimate have not been audited by the Air Force audit
service. 

--------------------
\28 The U.S.  Civil Service System's general schedule designates
positions by salary grade, from 1 to 15.  Salary increments between
grades are designated by steps.  Pay adjustments, known as locality
pay, also are made to compensate for disparities in Civil Service and
private sector salaries and wages for comparable positions. 

      SELECTING A MEASURE FOR THE
      COST-PER-CHILD ANALYSIS
      REQUIRED BALANCING ACCURACY
      AGAINST AVAILABILITY
------------------------------------------------------- Appendix I:2.2

Air Force and civilian child development centers counted the number
of children served in several ways, and only some of the information
was available by the age groups of the children served.  The Air
Force's semiannual report had information on hourly attendance; the
total number of hours children spent in the child development center
classrooms during the year; and the total number of children the
center had the physical space to house, also called the number of
"child care spaces." However, in the semiannual report, only the
information on the number of hours children spent in the child
development center classrooms was reported by age group.  The Cost,
Quality and Child Outcomes in Child Care Centers study data file had
information on the number of children enrolled and on the number of
children in attendance on the date of data collection.  Both measures
were available by the age groups of the children served.  Both
measures also were expressed as the number of full-time-equivalent
(FTE) children served--that is, the total number of children who
attended for the full day as well as children who attended for only
part of the day.  However, information on actual hourly attendance
and on the number of child care spaces in civilian centers was not
available in the Cost, Quality and Child Outcomes in Child Care
Centers study data file. 

Because the Air Force data on hourly attendance and the civilian data
on daily attendance were the most similar measures, and both were
available by age group, we selected them as measures for the number
of children served.  However, because the Cost, Quality and Child
Outcomes in Child Care Centers study's daily attendance measure was
counted only 1 day in the year, we converted it to a measure of the
total number of hours children in high-quality civilian centers spent
in the classroom during a year by assuming that the number of
children in attendance in civilian centers was the same on every day
of the year.  This assumption may have yielded a total that was
slightly more or slightly less than the actual number of children
served in civilian centers. 

      CURRENT TOTAL COST DATA ON
      CIVILIAN CHILD CARE CENTERS
      WERE NOT AVAILABLE
------------------------------------------------------- Appendix I:2.3

Because the most recent data available on the full cost of civilian
child care centers and on the wages of child care providers were
collected in 1992, we had to approximate the Cost, Quality and Child
Outcomes in Child Care Centers cost data, adjusted for inflation as
indicated by the Consumer Price Index (CPI).  While the Employment
Cost Index in the Economic Report of the President shows that
workforce wages increased at about the same rate as the CPI, this
measure would be less precise than the actual wage information
available for the Air Force child development center staff. 

      OUR AIR FORCE AND CIVILIAN
      COST ANALYSES DID NOT
      INCLUDE THE COSTS OF
      OVERSIGHT, INSPECTIONS, OR
      RESOURCE AND REFERRAL
      SERVICES
------------------------------------------------------- Appendix I:2.4

In the Air Force child development program, the Air Force administers
and incurs the cost for all program operations, including oversight,
inspection, resource and referral services for parents seeking child
care, and centers.  In the civilian child care center market,
however, entities other than child care centers pay for and perform
the regulatory and referral functions.  State governments monitor and
inspect child care centers, and both public and for-profit
organizations provide resource and referral services.  Civilian
resource and referral agencies provide a range of services--consumer
education, databases of services, child care supply studies, and
training for providers--but do not care for children.  Therefore, to
ensure comparability, we excluded the cost of regulatory and referral
activities from our analysis. 

      LABOR COST INFORMATION IN
      THE COST, QUALITY AND CHILD
      OUTCOMES DATA FILE WAS NOT
      AS DETAILED AS THAT IN THE
      SEMIANNUAL REPORT
------------------------------------------------------- Appendix I:2.5

Center directors provided wage and benefit information on all
caregiver staff but were given the option of providing either the
same level of detail for all additional staff positions or total
annual wages and benefits.  Some centers provided the detailed wage
and benefit information, but most centers did not.  It was not
possible, therefore, to separate the civilian center labor cost
information into direct and indirect labor costs for comparison with
the Air Force center labor cost data. 

We describe the measurement and analysis procedures that underlie all
of these issues in greater detail in the sections that follow. 

   PROCEDURES USED TO ESTIMATE AIR
   FORCE CHILD DEVELOPMENT CENTER
   COSTS
--------------------------------------------------------- Appendix I:3

To develop estimates of the cost of child care at Air Force child
development centers, we used fiscal year 1997 data from the Air Force
Semiannual Child Development Center Report concerning the number of
hours of child care provided at each center and wage rates for center
staff, as well as other data we obtained from a number of sources. 
In estimating the cost of providing care to specific age groups of
children at each center, we allocated costs other than direct labor
costs to each age group on the basis of that age group's share of the
total hours of care provided at the center. 

      DATA SOURCES USED TO
      ESTIMATE AIR FORCE COST
      COMPONENTS
------------------------------------------------------- Appendix I:3.1

The two Air Force Semiannual Child Development Center Reports filed
for each Air Force child development center in fiscal year 1997 were
the primary sources of the data used in our estimates of the cost of
care.  From these reports, we obtained information on the number of
hours of care for each age group of children during the 6-month
period covered by the report, the number of staff-hours devoted to
the care of each age group, the number of caregivers and other staff
at the center, the wage rates of those staff paid from
nonappropriated funds, and the grade levels of staff paid from
appropriated funds. 

A second data source was the questionnaire responses we received from
each Air Force child development center.  In October 1998, we sent
questionnaires to the directors of all child development centers (93)
located on the 67 Air Force bases in the United States.  By February
1999, all questionnaires had been returned.  The questionnaire
provided information on the distribution of caregiver staff by grade
level across the various age groups of children, the estimated number
of hours of volunteer help and the value of any donated items
received by the center during that fiscal year, the indoor square
footage of space occupied by the center, and the dollar amount of
expenditures made by the center during fiscal year 1997 for supplies. 

We collected the survey data to supplement the cost information
available in the child development centers' semiannual reports. 
Because some Air Force child development program coordinators
submitted a single semiannual report for two or more centers located
on their bases, we had only 75 semiannual reports.  Therefore, when
we entered the survey and semiannual report data into a single data
file, we combined the survey responses of bases that had submitted
one semiannual report but two or more survey questionnaires.  Thus,
for analysis purposes, the total number of Air Force child
development centers was 75. 

To determine the hourly wage rates to be used in estimating the cost
of labor performed at a center by employees paid by appropriated
funds, we referred to the 1997 GS pay schedule applicable to the
geographic location of the center.  To determine the per-square-foot
rental cost of indoor space, we used the General Services
Administration publication, Summary Report of Real Property Leased by
the United States Throughout the World as of September 30, 1996.  As
discussed below, we also used a few other data sources for specific
items of information in developing our cost estimates. 

      ESTIMATING PROCEDURES
------------------------------------------------------- Appendix I:3.2

Because the Air Force semiannual reports and our survey of child
development centers did not furnish information about the costs of
utilities, food, major equipment, occupancy, legal services, or
donations, we used supplementary data sources and estimating
procedures to develop cost estimates for these components.  In
general, we estimated direct labor costs for caring for each age
group of children from information contained in the semiannual report
that directly related caregiver staff-hours of work to each age
group. 

For all cost items other than direct labor, we had to make
assumptions regarding what portion of each should be considered
attributable to each of the various age groups of children served by
the center.  For almost every such cost item, we allocated the cost
proportionally according to each age group's share of the total
number of hours of child care provided by the center during fiscal
year 1997.  Thus, for example, if care for 3- to 5-year-olds
constituted 20 percent of a center's total child care hours, we
allocated 20 percent of that center's estimated annual cost of
utilities to the 3- to 5-year-old age group.  In three instances, we
varied this allocation procedure slightly.  In the cases of the costs
of cribs and playground equipment, it did not seem reasonable to
follow this procedure.  Only the youngest children would likely make
use of cribs, and, conversely, the youngest children would not make
use of playground equipment.  Therefore, we allocated the costs of
cribs to children 2 years of age or younger, and within that category
to each of these groups--6 weeks to 6 months, 6 to 12 months, and 12
to 24 months--on the basis of each group's total number of
child-hours.  In allocating the cost of playground equipment, we
excluded the 12-month-old and younger group and allocated the cost to
the remaining age groups according to each group's total number of
child-hours.  Similarly, when allocating food costs, we excluded the
12-month-old and younger group, since we believe that their parents
provide that group's food. 

The following paragraphs describe in more detail the cost components
included in our estimates and the method we used to estimate the
magnitude of each of them. 

         DIRECT LABOR
----------------------------------------------------- Appendix I:3.2.1

The term "direct labor" refers to the work of caregiver staff in
classrooms.  To estimate the cost of direct labor for each age group
of children at each center, we multiplied the total number of
staff-hours shown on the two fiscal year 1997 semiannual reports as
having been spent providing care to that age group by the relevant
hourly wage.  To determine the relevant hourly wage for each age
group's care, we needed three items of information:  the grade levels
of the staff members providing the care, the number of child-hours of
care provided by each grade level, and the hourly wage of each grade
level.  Since the semiannual reports did not associate grade levels
with the staff-hours shown, we used information from our center
questionnaire to estimate the grade level distribution of the
staff-hours. 

On our questionnaire, center directors were asked to indicate, for a
typical day, which grade levels of staff were assigned to each age
group of children and how many staff-hours of each grade level were
devoted to each age group of children.  We applied the staffing
pattern indicated in response to that question to the total number of
staff-hours shown in the semiannual reports for each age group.  For
example, if a center director's response indicated that on a typical
day one-half of the staff time devoted to the 3- to 5-year-old age
group was the time of a GS-4 staff person, we assumed that half of
the staff-hours shown on that center's semiannual report as devoted
to the 3- to 5-year-old age group were GS-4 staff-hours.  In those
instances in which a center questionnaire did not indicate which
grade levels of caregiver staff served a particular age group, we
applied to that age group the median wage rate among caregivers for
that age group at centers that did provide caregiver grade level
information for that age group. 

For caregiver staff paid from nonappropriated funds, we obtained the
hourly wage information for each grade level directly from the
semiannual report for the second half of fiscal year 1997.  For those
paid from appropriated funds, the GS employees, we used the GS hourly
wage rates for the employees' grade level shown in the 1997 GS pay
schedule for the locality pay area in which the center is located. 
By reviewing the Office of Personnel Management's Central Personnel
Data File, we determined for the job series in which most caregiver
staff are employed--Job Series 1702, Education and Training
Technician--and for each grade level in that series, the average step
of all employees at that grade level.  In our cost estimating
procedure, we assumed, for each GS-graded employee, that the employee
was at whatever the average step was for all employees at that grade
level.  Thus, for every GS-graded employee, we assumed that the
employee's hourly wage rate in fiscal year 1997 was the rate in the
center's geographic location applicable to the nationwide average
step for Job Series 1702 employees at the grade level of the employee
in question. 

To each hourly wage rate we added a factor to cover the cost of
fringe benefits.  Air Force officials informed us that for GS-graded
employees, fringe benefits equal 25 percent of salary, and for
employees paid from nonappropriated funds, fringe benefits equal 22
percent of wages.  Therefore, to account for benefit costs, we
multiplied all GS hourly wage rates by 1.25 and all nonappropriated
funds employees' wage rates by 1.22.  By applying the appropriate
estimated hourly wage (and benefit) rate to the number of staff-hours
reported for each age group of children, we arrived at an estimated
total direct labor cost for each age group. 

         INDIRECT LABOR
----------------------------------------------------- Appendix I:3.2.2

We considered as indirect labor all work performed by center
employees other than the direct caregivers.  Included in this
category is the work of salaried employees, such as the center
director, assistant directors, and training and curriculum
specialists, as well as hourly employees, such as administrative
staff and food service personnel, among others.  From the semiannual
reports for the second half of fiscal year 1997, we obtained the
number and pay grades of these salaried employees, and for the
nonsalaried employees, we obtained the number of hours worked each
week.  We then multiplied the number of hours per week worked by each
hourly employee, or estimated hours per week worked, by the
employee's reported, or assumed, hourly wage rate; we then multiplied
the resulting number by 52 (number of weeks) and added that total to
the annual salary of the salaried employees (including for both
categories of employees the cost of fringe benefits) to arrive at an
estimate of the center's total cost of indirect labor.  (Since
training and curriculum specialists also work with the family child
care providers and school-age caregivers, we allocated only one-third
of their salaries to Air Force center costs.) We then allocated a
portion of that total cost to each age group of children, on the
basis of its proportion of total child-care hours, to estimate the
indirect labor cost of providing care to each age group. 

         SUPPLIES
----------------------------------------------------- Appendix I:3.2.3

We obtained estimates of the cost of supplies directly from the
centers in their questionnaires.  We asked each center director to
estimate the cost of supplies purchased during fiscal year 1997. 
Questionnaires from a few centers did not provide an estimate of the
cost of supplies, so for each of those centers, we estimated supply
cost on the basis of the center's number of child-hours for the year. 
We developed the quantitative relationship between child-hours and
supply cost by examining the relationship between those two
quantities among the many centers that did provide estimates of the
cost of supplies. 

         FOOD
----------------------------------------------------- Appendix I:3.2.4

We estimated the centers' food costs on the basis of information
reported by each Air Force base's Appropriated Fund Expense Summary
for 1997 (7000.12 report).  As with other cost elements, we then
allocated this total cost across the various age groups of children,
with the exception of the 6-week to 12-month age group.  We excluded
this group because such young children do not consume the kind of
food purchased by the centers.  For a large number of centers, the
7000.12 report did not show the cost of food.  For those centers for
which food cost data were missing, we estimated food costs on the
basis of the number of meals served by the center.  We developed the
quantitative relationship between the number of meals served and food
cost by examining the relationship among centers that did provide
estimates of the cost of food. 

         UTILITIES
----------------------------------------------------- Appendix I:3.2.5

From the Air Force's Appropriated and Nonappropriated Fund Expense
Summary for 1997 we determined the total amount of utilities costs
incurred by all centers worldwide in fiscal year 1997.  We then
calculated the percentage of centers located outside the United
States and reduced the total amount of worldwide utilities costs by
that percentage, assuming that the remaining amount was attributable
to centers in the United States.  We then allocated that remaining
amount across all U.S.  centers on the basis of the total indoor
square footage of space used by each center.  Thus, for example, if
the indoor square footage of a given center constituted 5 percent of
the total indoor square footage of all domestic centers, that center
would be allocated 5 percent of the total estimated utilities costs
of all U.S.  centers. 

         MAJOR EQUIPMENT
----------------------------------------------------- Appendix I:3.2.6

We defined the cost of using major equipment as the 1-year allocation
of the acquisition cost of all capital equipment (equipment whose
purchase price exceeds $300).\29 Thus, the use of equipment cost
would be equal to the straight-line depreciation charge on equipment
that the center could take if it were a private sector firm. 
Therefore, it was necessary for us to determine, or assume, three
items of information:  the kinds and quantities of equipment a center
would use, the purchase price of each item of equipment, and the
expected useful life of each item of equipment.  To estimate the
kinds and quantities of equipment a center would use, we interviewed
the directors of two large centers, a medium-sized center, and a
small center who had reasonably complete information available on the
types of equipment that their centers used.\30 With a few exceptions,
we used the average of the two large centers' purchase costs for each
item of equipment as our estimate of the purchase cost for that same
item for all centers. 

To estimate the quantities of each item of equipment that a center
would use, we began by assuming that each center would use only one
of each item of kitchen, playground, classroom, and office equipment. 
For three items--cribs, coat racks, and shelving units--we assumed
that the quantity used by a center would vary according to the number
of children served by the center.  To estimate the quantity of each
of these items that each center would use, we began by setting as a
standard the equipment configuration used by one of the two large
centers from which we had obtained cost information.  We then
compared the total number of child-hours of care provided by each
center during fiscal year 1997 to the total for the center we were
using as our standard and calculated the percentage relationship
between the two.  We used this relationship as the basis for
estimating the quantity of each of the four items of equipment that
each center would use.  For example, if a given center had as its
total number of child care hours 10,000 hours and the "standard"
center had a total of 15,000 hours, we would assign to the center
two-thirds of the quantity of each item that the "standard" center
used.  If the standard center used 300 shelving units, we would
assume that the other center would use 200 shelving units.  The two
exceptions to this method involved cribs and outdoor playground
equipment.  Because we assumed that cribs would be used by only the
youngest children, we "assigned" cribs to centers on the basis of
total child-hours for only children 2 years of age or younger.  We
also assumed that outdoor playground equipment would not be used by
children under the age of 1. 

After estimating the quantity of each item of equipment a center
would use, we multiplied that quantity by the estimated unit purchase
cost of that item to arrive at an estimate of the total purchase cost
of that item.  We then divided that cost by the number of years of
expected useful life of the item to obtain an estimated annual usage
cost.  The measures of expected useful life we used for the various
items of equipment ranged from 5 to 15 years and were based on the
experience reported to us by the center directors we had interviewed
regarding the kinds of equipment used at their centers.  Finally,
since the purchase prices we used in our analysis were 1998 prices
and the centers acquired much of the equipment in previous years, we
converted our 1998 usage costs to costs expressed in an earlier
year's dollars in order to replicate as closely as possible the
purchase costs the centers had incurred when they acquired the
equipment.  For purposes of this conversion, we made an assumption
that each item of equipment being used at centers in fiscal year 1997
had at that time been in use for half its useful life.  Thus, for
example, if a category of equipment has a useful life of 10 years, we
assumed that each such item of equipment had been purchased 5 years
earlier, or in fiscal year 1993.  We thus deflated the 1998 purchase
price of all such items to their 1993 price.  We added the estimated
annual usage costs of all items of equipment in each center to
estimate that center's total annual equipment usage cost.  The total
cost of each center's use of equipment was then allocated across the
center's age groups of children. 

--------------------
\29 Major equipment included items in a child development center's
kitchen, office, classroom, and playground, such as refrigerators,
freezers, television cameras and monitoring systems, swings, and
outdoor playhouses. 

\30 We contacted centers on Air Force bases that reported having only
one child development center.  A center was considered small if, on
the semiannual report submitted for fiscal year 1997, the director
reported 89 or fewer full-day children in attendance, medium if the
director reported 90 to 124 full-day children in attendance, and
large if the director reported 125 or more full-day children in
attendance. 

         OCCUPANCY
----------------------------------------------------- Appendix I:3.2.7

The cost element that we refer to as occupancy represents the annual
cost of using the space occupied by each center.  We defined
occupancy cost as the annual rent that a center would be required to
pay if it were to lease its space on the commercial rental market. 
In the center questionnaire, we asked the center director to indicate
the total square footage of indoor space used for center activities,
including administrative activities.  We multiplied that number by
$5.93 to arrive at an estimated annual occupancy cost.  We chose the
$5.93-per-square-foot rate on the basis of the average cost per
square foot of space leased by the Air Force in 1996, as stated in
the 1999 General Services Administration's report on real property. 
That cost was $5.80.  Using the CPI to determine the inflation
factor, we raised the $5.80 cost to the $5.93 figure. 

         LEGAL SERVICES
----------------------------------------------------- Appendix I:3.2.8

We based our estimate of the annual cost of legal services provided
to each center on the opinion of officials of the Office of the Air
Force Judge Advocate General regarding the extent of provision of
such services.  Those officials estimated that, on average, about
1-1/2 hours per month of the time of a lieutenant colonel or a
colonel would be spent in providing such services.  We therefore
determined the midpoint between the lowest annual pay rate of the
lieutenant colonel and highest pay rate of the colonel and divided
that amount by the 2,080 hours in a federal work year to arrive at an
hourly rate.  We multiplied that figure by 1.5 to estimate a monthly
cost and then multiplied that cost by 12 to estimate the annual cost
of the legal services provided to each center.  We then allocated
that cost across the age groups of children receiving care at each
center. 

         DONATIONS
----------------------------------------------------- Appendix I:3.2.9

In the center questionnaire we asked each center director to estimate
the number of hours of volunteer help the center received during a
typical week in fiscal year 1997.  For each center, we multiplied the
number of reported hours of such help by 52 to estimate the total
number of hours of such work for the year.  We then multiplied that
total by the minimum wage rate of $5.15 an hour in effect in 1997 to
estimate the total value of volunteer labor donated to the center in
fiscal year 1997.  We then allocated that total figure across the
various age groups of children served by the center. 

In the center questionnaire we also asked each center director to
estimate the total value of donations of such items as toys and
supplies received by the center during fiscal year 1997.  We
allocated that total value across the various age groups of children. 

         OTHER COSTS
---------------------------------------------------- Appendix I:3.2.10

We did not attempt to estimate an imputed value for insurance or
taxes for the Air Force child development centers. 

   PROCEDURES USED TO ESTIMATE
   CIVILIAN CHILD CARE COSTS
--------------------------------------------------------- Appendix I:4

The Cost, Quality and Child Outcomes in Child Care Centers study data
file was the only source of data we used to develop estimates of the
cost of care at civilian child care centers.  The Cost, Quality and
Child Outcomes study team was an interdisciplinary group composed of
economists and child development psychologists from universities in
four states, led by the University of Colorado at Denver.  The team
selected the child care centers examined in the study using a
stratified random sample of approximately 100 centers in subregions
of each of four participating states, with approximately equal
representation of for-profit and nonprofit centers.  The four
participating states were California, Colorado, Connecticut, and
North Carolina. 

The data we used most extensively were from an interview with the
director of each child care center in the subsample.  The interview
was quite extensive and included all of the cost items. 

      SELECTION OF THE SUBSAMPLE
      OF HIGH-QUALITY CIVILIAN
      CENTERS
------------------------------------------------------- Appendix I:4.1

An important aspect of the Cost, Quality and Child Outcomes in Child
Care Centers study that made the data suitable for comparison with
the Air Force child development centers was a measure of the quality
of each civilian child care center.  Because DOD child development
centers are acknowledged to be of high quality, it was important to
compare the Air Force centers selected for the cost analysis only
with high-quality civilian centers. 

The study employed several measures of center quality, but two in
particular for centers with infants, toddlers, and preschoolers. 
These measures were obtained by direct observation.  Two members of
the study team randomly picked two rooms in the center and observed
the interaction in the rooms for 6 hours.  The first measure was the
ECER/ITER scale, a combined measure of the ECERS and ITERS.  The
studies' authors used these because they are well established global
measures of child care processes.  The second measure the authors
computed was a scale called INDEXECW, which is a weighted process
index scaled to ECERS and includes measures from the CIS and TIS.  We
selected our high-quality subsample by choosing centers that scored
4.5 or higher on either of these two 7-point scales.  The resulting
subsample included 150 centers or 37 percent of the Cost, Quality and
Child Outcomes study population. 

      ESTIMATING PROCEDURES
------------------------------------------------------- Appendix I:4.2

The Cost, Quality and Child Outcomes study team collected extensive
information on the costs of operating a civilian child care center. 
This information allowed us to compare costs with those of Air Force
child development centers.  Information on the following components
of cost was obtained during the interview of the director.  Directors
were asked to provide records of the last fiscal year's expenses for
review by the study team.  The following cost components were
collected as total annual costs for the center, per year:  (1) wages,
(2) nonwage benefits, (3) staff education/training costs, (4)
subcontractor costs, (5) occupancy cash costs, (6) food service
costs, (7) insurance costs, (8) other operating costs, (9) overhead
costs, and (10) the subsidy supplied.  The subsidy is the estimated
value of donated goods and services, such as volunteer labor, donated
food, building space, and utilities.  We used these cost components
to estimate the overall cost per child and cost per child by age
group for the high-quality civilian centers.  Regarding labor costs,
although a few centers provided separate direct and indirect labor
costs, most centers in the Cost, Quality and Child Outcomes in Child
Care Centers study did not.  Therefore, we developed our own
estimates of the portion of total labor costs that should be
considered direct labor and the portion that should be considered
indirect labor.  For those few centers that provided separate amounts
for direct and indirect labor costs, we calculated the overall
proportions of total labor costs attributable to direct and indirect
labor costs, which were about 82 percent and 18 percent,
respectively.  We then applied these proportions to total labor costs
for all centers. 

The cost data were collected in the spring of 1993 but were requested
for the latest fiscal year.  We assumed the latest fiscal year would
be 1992 and adjusted the civilian costs from 1992 to 1997 by using
the CPI. 

      ALLOCATING COSTS BY
      CHILD-HOURS
------------------------------------------------------- Appendix I:4.3

As noted, the study provided annual costs for the entire center, not
costs by age group.  Because we needed to compare the civilian with
the military child care costs by age group, we needed to allocate
center costs by the proportion that each age group's child-hours
constituted of the center total.  Once we calculated the number of
child-hours in each age group, we knew the proportion each age group
represented of the whole. 

      CHILD-HOURS
------------------------------------------------------- Appendix I:4.4

During the director's interview, the director was asked what age
group each room in the center contained and how many FTE children
were there on that particular day.  These responses were used to
determine whether each room had infants (0 to 18 months), toddlers
(19 to 30 months), preschool-aged children (2 to 5 years),
kindergarten-aged children (5 years), or school-aged children (6
years and older).\31 We used the reported age group in each room to
add up the FTE children for different age groups across the 150 high
quality-centers.  The total number of FTE children for the center for
the year is calculated from the number there on the day of the
interview.  First, we multiplied the number of children present on
the day of the interview by eight, resulting in the number of
child-hours per day.  We then multiplied the daily number of hours by
five (the number of days in a week that most centers were open) to
get weekly hours.  Next, we multiplied the weekly hours by 52 (weeks
per year).  If a center closed for the summer, we subtracted the
number of months the center was closed times 4.3 (average number of
weeks in a month) from the total for a yearly figure.  We then
allocated the costs on the basis of the percentage of total
child-hours each age group had. 

--------------------
\31 The ages of children in the age group categories are approximate. 
The Cost, Quality and Child Outcomes study team learned during data
collection that children in many centers were assigned to groups on
the basis of their developmental level, rather than their
chronological age.  Therefore, because very young children develop at
different rates, it was possible that children as old as 18 months
were placed in infant rooms, and children as young as 2-1/2 might be
placed in preschool rooms. 

      CAREGIVER COMPENSATION RATE
------------------------------------------------------- Appendix I:4.5

The Cost, Quality and Child Outcomes in Child Care Centers study
included the hourly wage rate for many of the caregivers and aides,
but not all.  The rest of the compensation information consisted of
annual salaries.  We divided the annual salary by 52 weeks times the
number of hours worked per week (minus the number weeks the center
was closed) to achieve an hourly rate.  We then calculated total
compensation by adding employee benefits, which were equivalent to
19.73 percent of the wage rate, to the wage. 

      CHILD-TO-STAFF RATIOS
------------------------------------------------------- Appendix I:4.6

The Cost, Quality and Child Outcomes in Child Care Centers study data
file had information that allowed us to compute ratios of the number
of children per caregiver in the centers' classrooms, known as the
"child-to-staff ratio." One segment of the data file contained
information on the number of hours a week that each caregiver and
aide worked.  First, we divided the total number of hours worked by
all of the caregivers in our subsample by the total number of FTE
caregivers.  This gave us the average number of hours worked.  We
then multiplied that figure by 52, and subtracted the time the center
was closed.  Because we already had child-hours, this additional
information on caregiver-hours allowed us to calculate a ratio of
child-hours to caregiver-hours for each age group. 

   FACTORS AFFECTING AIR FORCE
   COSTS
--------------------------------------------------------- Appendix I:5

To further understand how much the age distribution of the centers'
children affected the overall Air Force cost per child-hour, we
estimated what the Air Force centers' cost would be using the age
distribution of the civilian centers' children.  We reestimated the
Air Force overall cost per child-hour, substituting the age
distribution of the civilian centers' children for the age
distribution of Air Force children in the total cost estimate.  This
estimate allowed us to calculate the dollar value of the difference
that the age distribution of the children contributed to the overall
cost per child-hour. 

CHILD-TO-STAFF RATIOS WITHIN GROUP
SIZE RECOMMENDED BY NAEYC
========================================================== Appendix II

                                                    Group size
                         ----------------------------------------------------------------
Age of children             6     8    10    12    14    16    18    20    22    24    28
-----------------------  ----  ----  ----  ----  ----  ----  ----  ----  ----  ----  ----
Infants (birth-12         3:1   4:1
 months)
Toddlers (12-24 months)   3:1   4:1   5:1   4:1
2-year-olds (24-30              4:1   5:1   6:1
 months)
2-1/2-year-olds (30-36                5:1   6:1   7:1
 months)
3-year-olds                                       7:1   8:1   9:1  10:1
4-year-olds                                             8:1   9:1  10:1
5-year-olds                                             8:1   9:1  10:1
6-to 8-year-olds                                                   10:1  11:1  12:1
9-to 12-year-olds                                                              12:1  14:1
-----------------------------------------------------------------------------------------
Notes:  Smaller group sizes and lower child-to-staff ratios have been
found to be strong predictors of compliance with indicators of
quality, such as positive interactions among staff and children and
developmentally appropriate curriculum.  Variations in group sizes
and ratios are acceptable in cases in which programs demonstrate a
very high level of compliance with criteria for interactions,
curriculum, staff qualifications, health and safety, and physical
environment. 

Source:  NAEYC. 

FISCAL YEAR 1997 PER-CHILD-HOUR
COST COMPONENTS FOR AIR FORCE
CHILD DEVELOPMENT CENTERS, BY AGE
GROUP
========================================================= Appendix III

                                                                                                                           Part-day
                                        0-6 months           6-12 months   12-24 months   24-36 months    3-5 years       preschool       6 and over
                                --------------------------  -------------  ------------  --------------  ------------  ----------------  ------------
                                                               Cost  % of    Cost  % of            % of    Cost  % of                      Cost  % of
                                    Cost per                    per  tota     per  tota  Cost per  tota     per  tota  Cost per    % of     per  tota
Cost components                         hour    % of total     hour     l    hour     l      hour     l    hour     l      hour   total    hour     l
------------------------------  ------------  ------------  -------  ----  ------  ----  --------  ----  ------  ----  --------  ------  ------  ----
Direct labor                           $3.73          68.8    $3.75  69.5   $2.84  60.2     $2.12  53.3   $1.38  42.7     $1.41    42.4   $1.34  45.6
Indirect labor                          0.90          16.6     0.87  16.2    0.91  19.4      0.90  22.7    0.88  27.3      0.94    28.3    0.70  24.0
Supplies                                0.28           5.2     0.27   5.1    0.27   5.8      0.27   6.8    0.27   8.5      0.23     7.0    0.19   6.4
Utilities                               0.07           1.4     0.07   1.4    0.07   1.5      0.07   1.8    0.07   2.3      0.08     2.3    0.08   2.8
Food                                       0             0        0     0    0.21   4.5      0.21   5.3    0.21   6.3      0.21     6.3    0.18   6.1
Equipment                               0.04           0.7     0.04   0.8    0.02   0.4      0.03   0.7    0.03   0.9      0.03     1.0    0.02   0.7
Space                                   0.38           7.1     0.38   7.0    0.37   7.9      0.37   9.2    0.38  11.7      0.39    11.8    0.42  14.2
Legal services                             0             0        0     0       0     0         0    .1       0    .1         0      .1       0    .1
Volunteers                              0.01            .2     0.01   0.2    0.01   0.2      0.01   0.2    0.01   0.3      0.03     0.8    0.01   0.2
Donations                                  0             0        0     0       0     0         0    .1       0    .1         0       0       0     0
=====================================================================================================================================================
Total                                  $5.43           100    $5.40   100   $4.72   100     $3.97   100   $3.23   100     $3.33     100   $2.94   100
-----------------------------------------------------------------------------------------------------------------------------------------------------
Note:  Because of rounding, individual percentages in a column do not
always equal 100 percent and cost components do not always equal
totals. 

Source:  GAO analysis of Air Force center data. 

(See figure in printed edition.)Appendix IV
COMMENTS FROM THE DEPARTMENT OF
DEFENSE
========================================================= Appendix III

(See figure in printed edition.)

GAO CONTACTS AND STAFF
ACKNOWLEDGMENTS
=========================================================== Appendix V

GAO CONTACTS

Karen Whiten, Assistant Director, (202) 512-7291
Sara Edmondson, Evaluator-in-Charge, (202) 512-8516

STAFF ACKNOWLEDGMENTS

In addition to those named above, the following individuals made
important contributions to this report:  Elizabeth Morrison was
responsible for determining the Air Force cost estimates and
analyzing DOD's child development program; James Wright provided
guidance on study design, measurement, and data analysis; John Smale,
Jr., provided survey and data analysis support; Joan Vogel created
the data files and provided the computer programming for estimating
the Air Force and civilian costs; and Janet Mascia and Alicia Cackley
provided technical guidance on child care quality and cost
measurement issues. 

*** End of document. ***