Head Start: Better Data and Processes Needed to Monitor 	 
Underenrollment (04-DEC-03, GAO-04-17). 			 
                                                                 
Head Start, created in 1965, is designed to prepare low-income	 
preschool children for school by providing a comprehensive set of
early child development services primarily through communitybased
organizations. Over the last decade there have been a number of  
changes in Head Start's operating environment, including a	 
decrease in the number of poor children; an increase in the	 
number, size, and scope of other federal and state early	 
childhood programs; and an expansion in Head Start spending and  
enrollment. Given this environment, GAO was asked to determine	 
(1) what is known about the extent to which Head Start programs  
are underenrolled, (2) ACF regional officials' and Head Start	 
grantees' views on what factors contribute to underenrollment,	 
and (3) what actions ACF and grantees have taken to address	 
underenrollment.						 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-17						        
    ACCNO:   A08969						        
  TITLE:     Head Start: Better Data and Processes Needed to Monitor  
Underenrollment 						 
     DATE:   12/04/2003 
  SUBJECT:   Children						 
	     Preschool education				 
	     Education program evaluation			 
	     Regional development programs			 
	     Monitoring 					 
	     Head Start Program 				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO 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.                                         **
**                                                              **
******************************************************************
GAO-04-17

United States General Accounting Office

GAO

                       Report to Congressional Requesters

December 2003

HEAD START

          Better Data and Processes Needed to Monitor Underenrollment

GAO-04-17

Highlights of GAO-04-17, a report to Congressional Requesters

Head Start, created in 1965, is designed to prepare low-income preschool
children for school by providing a comprehensive set of early child
development services primarily through communitybased organizations. Over
the last decade there have been a number of changes in Head Start's
operating environment, including a decrease in the number of poor
children; an increase in the number, size, and scope of other federal and
state early childhood programs; and an expansion in Head Start spending
and enrollment. Given this environment, GAO was asked to determine (1)
what is known about the extent to which Head Start programs are
underenrolled, (2) ACF regional officials' and Head Start grantees' views
on what factors contribute to underenrollment, and (3) what actions ACF
and grantees have taken to address underenrollment.

GAO recommends that the Secretary of HHS direct ACF to (1) ensure the
accuracy of national enrollment data, (2) develop a standard criterion for
regional offices to use in identifying grantees whose underenrollment
merits action, (3) develop an additional enrollment measure that takes
into consideration the different levels of service provided by full-day
and part-day programs, and (4) develop a more systematic process for
regional offices to collect reliable enrollment data during the program
year.

www.gao.gov/cgi-bin/getrpt?GAO-04-17.

To view the full product, including the scope and methodology, click on
the link above. For more information, contact Marnie S. Shaul at (202)
512-7215 or [email protected].

December 2003

HEAD START

Better Data and Processes Needed to Monitor Underenrollment

The extent to which Head Start programs have enrolled fewer children than
they are funded to serve is unknown because the Administration for
Children and Families (ACF) does not collect accurate national data and
does not monitor underenrollment in a uniform or timely manner. While some
modest fluctuations in enrollment are to be expected, regional offices had
differing definitions of unacceptable underenrollment, and the approaches
they used to identify it were either not timely or not systematic. The
regional offices identified a total of about 7 percent of grantees as
unacceptably underenrolled in 2001-02, significantly less than the
percentage of grantees reporting enrollment ratios below 100 and 95
percent on ACF's survey of grantees (see chart below). As a result of
differences in regional definitions of what constitutes an unacceptable
level of underenrollment, grantees with similar levels of underenrollment
may be treated differently across regions.

ACF regional officials and officials of underenrolled Head Start grantees
often cited a mixture of factors that made it difficult to achieve full
enrollment, including increased parental demand for full-day child care, a
decrease in the number of eligible children, facilities-related problems,
and more parents seeking openings with other sponsors of early education
and care.

ACF national and regional offices and grantees all report taking action to
address underenrollment through the issuance of guidance, increased
monitoring by regional offices, and more aggressive outreach attempts by
grantees. The ACF national office issued a memo in April 2003 that
instructed regional offices to address underenrollment with a variety of
measures depending on its causes. While this guidance was clear on the
actions to be taken, it lacked clear criteria for prioritizing grantees
for corrective actions. Also, while many grantees we spoke with had taken
steps to address underenrollment, some told us of their concern to
maintain total funded enrollment levels, even as they were converting
unfilled part-day openings to full-day. While 18 of the 25 grantees we
contacted had made progress toward full enrollment, others cited
continuing problems.

Percentage of Underenrolled Grantees Using Three Different Definitions

                           Less than 100% enrolled 57

Less than 95% enrolled

Regionally identified as "unacceptable"

0 102030405060

Percentage

Source: GAO analysis of ACF regional office survey responses and Program
Information Report data.

Contents

     Letter                                                                 1 
                                    Results in Brief                        2 
                                       Background                           4 
              The Extent to Which Head Start Programs Are Underenrolled is 
                                       Not Known                           11 
              Regional and Grantee Officials Often Cited Combinations of   
                       Factors as Responsible for Underenrollment          18 
                ACF and Grantees Use a Variety of Approaches to Address    
                                    Underenrollment                        22 
                                      Conclusions                          28 
                          Recommendations for Executive Action             29 
                                    Agency Comments                        30 
Appendix I                    Scope And Methodology                     

Appendix II 	Factors That the ACF Regions and the American Indian-Alaska
Native Program Branch Believed Contributed to Underenrollment to a Major
or Moderate Extent

Appendix III 	Factors that Grantees Believed Contributed to Their Head
Start Programs' Underenrollment

Appendix IV 	Comments from the U.S. Department of Health and Human
Services

Appendix V GAO Contacts and Staff Acknowledgments 37

GAO Contacts 37 Acknowledgments 37

Related GAO Products

38

Tables

Table 1: ACF Regional Thresholds for Unacceptable Levels of
Underenrollment

Table 2: Number of ACF Regions Relying on Various Methods to Oversee Head
Start Grantee and Delegate Agency Enrollment Levels

Table 3: Percentage of Head Start Grantees, by Region, Reported as
Unacceptably Underenrolled by ACF Regions with Unacceptable
Underenrollment Thresholds, Compared with PIR Data at the Same Thresholds

Table 4: Comparison of Region II and III Grantees Identified as
Underenrolled

Table 5: April 2003 National Head Start Guidance to ACF Regions

Table 6: Actions by ACF Regions toward Underenrolled Grantees

Table 7: Actions Taken by Interviewed Grantees to Address Underenrollment

Table 8: Factors Cited By ACF Regions as Contributing to Underenrollment
to a Major or Moderate Extent, during Program Years 2001-02 and 2002-03

Table 9: Factors Cited by 25 Grantees as Contributing to Head Start
Program Underenrollment, during Program Years 2001-02 and 2002-03

                                     13 14

17

18 24 25

27

33

34

Figures

Figure 1: ACF Regions Figure 2: Distribution of Head Start Services
between Full-day and Part-day Figure 3: Head Start Enrollment Compared
with Children Under Age 6 Living in Poverty (1992-2002) Figure 4: Growth
in Federal Investment in Child Care, Fiscal Years 1997 through 2002 in
Nominal Dollars

Figure 5: Comparison of Enrollment Ratios as reported in 2001-02 PIR
Survey with Unacceptably Underenrolled Grantees as Reported by ACF Regions

                                                                     5 6 8 10

15

Abbreviations

ACF Administration for Children and Families
CCDF Child Care Development Fund
HHS Department of Health and Human Services

PIR Program Information Report
SSI Supplemental Security Income
TANF Temporary Assistance for Needy Families

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States General Accounting Office Washington, DC 20548

December 4, 2003

The Honorable George Miller
Ranking Minority Member
Committee on Education and the Workforce
House of Representatives

The Honorable Dale E. Kildee
House of Representatives

The Honorable Adam Schiff
House of Representatives

Head Start is the largest federal early childhood program, funded at about
$6.7 billion in fiscal year 2003. Created in 1965, Head Start is designed
to
prepare poor children for school by providing a comprehensive set of
developmental services. The Administration for Children and Families
(ACF) of the Department of Health and Human Services (HHS)
administers the program. Over the last decade, the Head Start program has
expanded substantially. Between fiscal years 1990 and 2002,
appropriations for Head Start quadrupled from $1.6 billion to over
$6.5 billion, and the number of children served increased by 69 percent
from about 540,000 to over 910,000.

Over the past decade, significant changes to Head Start's environment may
have created challenges for some Head Start grantees when they tried to
find children to fill funded slots. For example, in the 1990s there was a
decline in welfare caseloads following welfare reform and a decline in the
number of children living in poverty, which may have decreased the
number of children eligible for Head Start. At the same time, the
expansion of other federal and state early childhood programs may have
increased child care options available to Head Start-eligible families.
Consequently, it is possible that federally funded Head Start slots in
some
areas remain unfilled even while eligible children elsewhere remain on
waiting lists. Given these potential challenges, and in anticipation of
Head
Start's reauthorization, you asked that we determine the extent to which
Head Start grantees were underenrolled and that we identify potential
causes of underenrollment. As agreed with your offices, our review
addresses: (1) what is known about the extent to which Head Start
programs are underenrolled, (2) what factors may have contributed to

underenrollment, and (3) what actions ACF and grantees have taken to
address underenrollment.

To determine what is known about the extent to which Head Start programs
are underenrolled, we attempted to verify the accuracy of national
enrollment data, interviewed ACF headquarters officials, and reviewed
federal guidance and regulations on enrollment. Because we determined that
national enrollment data were not reliable and because the regional
offices have primary responsibility for identifying and addressing
underenrollment, we surveyed all 10 ACF regional offices and the American
Indian-Alaska Native Program Branch.1 We asked them to identify the
threshold below which they consider underenrollment to be unacceptable and
to identify grantees with enrollment levels beneath this threshold. To
gather further details on the process by which regions identify and
address underenrollment, we interviewed regional officials in 3
regions-region III (covering Delaware, the District of Columbia, Maryland,
Pennsylvania, Virginia, and West Virginia), region V (Illinois, Indiana,
Michigan, Minnesota, Ohio, and Wisconsin), and region IX (Arizona,
California, Hawaii, Nevada, and the Pacific Insular Areas)- selected on
the basis of geographical representation and the number of underenrolled
grantees they reported to us. To determine what factors ACF officials and
Head Start grantees believed contributed to underenrollment and to
identify actions they took to address it, we surveyed all 10 ACF regional
offices and the American Indian-Alaska Native Program Branch office,
selected and interviewed 25 grantees identified by regional offices as
unacceptably underenrolled, and conducted site visits to 3 regional
offices listed above. We performed our work between May and October 2003
in accordance with generally accepted government accounting standards.
Appendix I further describes our scope and methodology.

The extent to which Head Start programs are underenrolled is unknown
because ACF does not collect accurate national data and its regional
offices do not monitor grantee enrollment in a uniform or timely manner.
The agency surveys grantees annually to determine national enrollment
levels, but we found these data contained many inaccuracies and were

1We did not survey the Migrant and Seasonal Program Branch due to the
program's distinct seasonal operating schedule. We generally will refer to
the 10 regions and 1 program branch we surveyed as "regional offices" or
"regions" throughout the report.

  Results in Brief

unavailable during the current program year for regional use in monitoring
grantees. Consequently, we could not determine the extent to which Head
Start grantees had enrollments below 100 percent of funded enrollment,
which is how HHS' regulations define underenrollment. We found that three
regional offices did not set a threshold below which they view
underenrollment as unacceptable, while the other regions use thresholds
ranging from anything below 100 percent to enrollment below 74 percent. We
also found that the approaches regions used to identify unacceptable
levels of underenrollment, such as visits to grantees and reviews of grant
re-funding applications and grantee audits, were either not timely or did
not systematically address underenrollment. Using varying thresholds, the
regional offices identified a total of 170 grantees as unacceptably
underenrolled in program year 2001-02, or about 7 percent of all grantees.
By contrast, using the regulatory definition of underenrollment, survey
data indicated that as many as half or more of the grantees could be
underenrolled. As a result of differences in regional thresholds for what
constitutes an unacceptable level of underenrollment, grantees with
similar levels of underenrollment may be treated differently across
regions.

ACF regional officials and officials of underenrolled Head Start grantees
often cited combinations of factors that made it difficult to achieve
acceptable levels of enrollment, including increasing parental demand for
full-day child care and decreasing numbers of eligible children. Providing
full-day care was said to be more expensive than providing part-day care
because it would require more facility space and staff per child. To make
full-time slots available, many grantees we spoke with were attempting to
expand their facilities or partner with other programs. However, 14 of the
25 grantees we interviewed reported having difficulty acquiring and
developing adequate facilities. Underenrolled grantees and regional
officials also said that underenrollment occurred because parents were
increasingly seeking services from other early education and child care
programs, some of which subsidized care provided by relatives. Other
contributing factors were less frequently cited, such as eligible families
moving from the service area, language and cultural differences between
children's families and program staff, and weak or inadequate outreach
efforts by grantees to locate eligible families.

ACF national and regional offices and grantees all report taking actions
to address underenrollment, such as issuing guidance, increasing
monitoring, and more aggressively trying to recruit participants. The ACF
national office issued a memorandum in April 2003 that instructed regional
offices to address underenrollment in particular ways depending on its
underlying

cause. However, the guidance lacked clear criteria for prioritizing which
grantees should be subject to corrective action based upon their level of
underenrollment. According to regional officials, the actions they most
frequently took to address underenrollment were to monitor enrollment
levels, track improvement efforts, and provide training and technical
assistance. Many grantees we spoke with have also taken steps to address
underenrollment, such as increasing their outreach efforts, seeking
partners to help them provide more full-day service, or increasing the
availability of full-day slots. Furthermore, some grantees told us of
their concern to maintain total funded enrollment levels, even as they
were converting unfilled part-day openings to full day. Consequently, in
some instances, grantees attempting to convert part-day slots to full-day
slots said they had to expand facilities or find other child care partners
in order to serve the same number of children. While 18 of the 25 grantees
we contacted had made progress toward achieving full enrollment, others
cited continuing problems.

To improve ACF's ability to identify and address underenrollment in a more
systematic and timely manner, we are making recommendations that the
agency improve the quality of enrollment data and establish more uniform
criteria and procedures for identifying and addressing underenrollment.

Head Start was designed to help break the cycle of poverty by providing
comprehensive educational, social, health, nutritional, and psychological
services to low-income children. Head Start is authorized to serve
children at any age prior to compulsory school attendance. Originally, the
program was aimed at 3- to 5-year-olds. A companion program begun in 1994,
Early Head Start, made these services available to children from birth to
3 years of age as well as to pregnant women. Head Start and Early Head
Start programs are administered by ACF, which funds and monitors more than
1,500 grantees through its 10 regional and 2 branch offices. (See fig. 1)
ACF's national office has responsibility for overseeing and providing
guidance to the regional offices, as well as for administering and
collecting annual survey data from grantees.

  Background

Figure 1: ACF Regions

Source: ACF.

Head Start grantees include community action agencies, school systems,
for-profit and nonprofit organizations, other government agencies, and
tribal governments or associations. Also, many Head Start grantees provide
services by subcontracting with other organizations, known as delegate
agencies. In fiscal year 2002, Head Start grantees served more than
912,000 children, a 69 percent increase over the number of children served
in 1990. Head Start has traditionally been a part-day, part-year program,
but currently serves more children on a full-day basis, which is defined
as 6 hours or more a day, than on a part-day basis. Approximately 47
percent of children served by Head Start were enrolled in a centerbased
full-day program for 6 hours or more a day.2 Less than 20 percent of
children enrolled in Head Start receive 8 hours or more of center-based
services a day. As of 2001-02, about 44 percent of Head Start children
were

2Center-based programs are those where services are provided to children
primarily in classroom settings.

enrolled in a part-day center-based program. Figure 2 shows the
percentages of Head Start services provided on a full-day or part-day
basis.

Figure 2: Distribution of Head Start Services between Full-day and
Part-day

Other

Center-based full-day greater than or equal to 8 hours per day

Center-based full-day between 6 and 8 hours per day

Center-based part-day less than 6 hours per day Source: GAO analysis of
2001-02 Program Information Report Data.

Head Start funds are allotted among the states based on their 1998
allocation and, for funds exceeding that amount, by formula based on the
number of children in each state under the age of 5 from families whose
income is below the federal poverty level.3 Head Start grantees are
required to provide at least 20 percent of annual program funding, which
can include in-kind contributions, such as facilities for holding classes.
During the award process, Head Start grantees receive from ACF regional
officials their level of funded enrollment-the number of children the
grantee is to serve.

Head Start regulations require that at least 90 percent of the children
enrolled in Head Start come from families with incomes at or below the
federal poverty guidelines, from families receiving public assistance, or
from families caring for a foster child. While the poverty guidelines are

3For 2003, the federal poverty line for a family of four was $18,400
within the 48 contiguous states and the District of Columbia. In Alaska
and Hawaii, the guidelines were $23,000 and $21,160, respectively. The
poverty guidelines are updated periodically in the Federal Register, by
the U.S. Department of Health and Human Services under the authority of 42
U.S.C. 9902(2).

firm, grantees have some flexibility in determining income eligibility.
For example, grantees can use the 12 months prior to the month the family
applied to Head Start or the previous calendar year as a basis for
determining income eligibility. Also, once a family is determined to be
eligible in 1 program year, it is considered eligible for the subsequent
program year, for a total of 2 years. Additionally, families that
participate in the Temporary Assistance for Needy Families program (TANF)
or the Supplemental Security Income program (SSI)4 or that care for a
foster child are eligible for Head Start services even when family income
exceeds the poverty guidelines. Grantees may fill up to 10 percent of
their slots with children from families that exceed the low-income
guidelines.

An enrollment occurs when a Head Start program officially accepts a child
and completes all necessary steps to begin providing services. If a child
is chronically absent and the grantee cannot serve the child in another
way, the child's slot is considered vacant. Once a slot is vacant, the
grantee generally must fill it within 30 days to be considered fully
enrolled. Consequently, actual enrollments can fluctuate somewhat
throughout funding periods. Head Start regulations require grantees to
track program attendance on a daily basis. However, grantees are asked to
annually report enrollment levels for any 2 months they choose as part of
ACF's annual Program Information Report (PIR) survey. ACF regions may
require grantees to report enrollment data more frequently.

Head Start regulations require grantees to maintain enrollment at 100
percent of the funded level and regional offices have primary
responsibility for identifying and addressing underenrollment.5 However,
as a practical matter, not all grantees are able to continually sustain
enrollment at the fully funded level. Underenrollment can occur for a
variety of reasons and can vary from month to month in a given program.
Therefore, before deciding that underenrollment is unacceptable and taking
action, the regions take into consideration a variety of factors about
underenrollment, including its level and duration, its causes, and the
actions taken by grantees to address it.

4Under TANF, the federal government provides grant funds to states,
territories, and tribes for their programs to assist needy families with
children. SSI pays monthly benefits to people who are age 65 or older or
blind or have a disability and who do not own much or have much income.

5Enrollment ratio is the ratio of actual enrollment to funded enrollment;
therefore, a grantee with 100 funded enrollment slots and an actual
enrollment of 95 has an enrollment ratio of 0.95, or 95 percent.

The number of children eligible for Head Start services on the basis of
being below the poverty line has decreased over the last decade, falling
from over 6 million children in 1992 to just over 4 million in 2000. By
2002, the number of children under age 6 living in poverty had increased
to nearly 4.3 million. Over the same period, Head Start enrollment has
increased to over 910,000 children-a level that is significantly below the
number of children living in poverty. (See fig. 3.) However, it should be
noted that Head Start predominately serves children ages 3 and 4, who make
up only a portion of all children under 6 living in poverty.

Figure 3: Head Start Enrollment Compared with Children Under Age 6 Living
in Poverty (1992-2002)

Number of children

7,000,000

6,000,000

5,000,000

4,000,000

3,000,000

2,000,000

1,000,000

0 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Years

Head Start enrollment

Children under age 6 and in poverty

Source: Poverty data are taken from the Census Bureau's Current Population
Survey, Annual Social and Economic Supplements. Enrollment figures are
from ACF's PIR data.

Note: Data on number of children under age 6 living in poverty includes
only those residing with a

relative.

In addition, during the 1990s, the number of other federal and state
programs offering services to low-income children increased substantially.
For example, welfare reform in 1996 greatly expanded the Child Care
Development Fund (CCDF) and also allowed TANF funds to be used for

child care.6 For fiscal years 1997 through 2002, these programs increased
their investment in children; CCDF spending increased from $2.5 billion to
$6.4 billion and TANF spending on child care increased from $13 million to
$1.6 billion.7 (See fig. 4.) On the state level, one study cited by the
Congressional Research Service (CRS) found that state spending on
prekindergarten programs increased from about $700 million in 1991-92 to
about $1.7 billion in 1998-99.8 Over the same period, the number of
children served by these programs has increased from 290,000 to 725,000.

6Other federal programs that support early childhood education for
children under 5 years include Special Education Preschool grants under
the Individuals with Disabilities Education Act, preschool programs under
Title I, and Even Start. See Education and Care: Head Start Key Among
Array of Early Childhood Programs, but National Research on Effectiveness
Not Completed, GAO-03-840T (Washington D.C.: July 22, 2003).

7CCDF child care funds can be used for families with children up to age
13. In 2000, we reported that an estimated 70 percent of CCDF funds and
7.5 percent of TANF funds were used for child care for children under age
5 in fiscal year 1999. See Early Education and Care: Overlap Indicates
Need to Assess Crosscutting Programs. GAO/HEHS-00-78. Washington D.C.:
April 28, 2000.

8See Blank, Helen, with Karen Schulman and Danielle Ewin. Seeds of
Success, State Prekindergarten Initiatives, 1998-99. Children's Defense
Fund, September 1999 as cited in Early Childhood Education: Federal Policy
Issues, CRS, (Washington, D.C.: January 27, 2003).

Figure 4: Growth in Federal Investment in Child Care, Fiscal Years 1997
through 2002 in Nominal Dollars

Dollars in millions 8,000

7,000

6,000

5,000

4,000

3,000

2,000

1,000

0 1997 1998 1999 2000 2001 2003 Years

TANF

CCDF

Sources: CRS Report, Child Care: Funding and Spending under Federal Block
Grants, March 19, 2002 and ACF budget figures.

Note: CCDF amounts include dollars states transferred from their TANF
programs to CCDF as allowed under the Personal Responsibility and Work
Opportunity Reconciliation Act of 1996. The amounts shown for TANF include
only those TANF funds expended for child care.

Expanding federal and state early childhood programs has increased the
need for coordination to better ensure that services are provided in a
complementary fashion. One way that Head Start encourages coordination is
by requiring all grantees to periodically prepare community assessments
that analyze trends in the number of eligible children in their
jurisdictions and assess the other early childhood services provided in
the area. While grantees are not required to coordinate with other service
providers, ACF has issued guidance encouraging grantees to coordinate with
other providers in order to provide more full-day services. Also, in
recent years, ACF has used some Head Start expansion money to build
partnerships with child care providers to deliver full-day, full-year
services. As another way to increase coordination, HHS has been authorized
since 1998 to provide additional funds to states to encourage such
collaboration.

  The Extent to Which Head Start Programs Are Underenrolled is Not Known

The extent to which Head Start programs are underenrolled is unknown
because ACF does not collect accurate national data and it does not
monitor grantee enrollment in a uniform or timely way. Specifically,
national enrollment data contain many inaccuracies and regional offices
use a variety of thresholds to define "unacceptable" levels of
underenrollment.9 Additional approaches used by the regions to identify
underenrollment do not systematically address underenrollment or provide
timely information. Using varying thresholds, the regional offices
identified 170 grantees as unacceptably underenrolled in program year
2001-02, or about 7 percent of all grantees. By contrast, the agency's
annual survey data indicated that as many as half or more of the grantees
were enrolled at less than 100 percent-the enrollment level grantees are
required to maintain under Head Start regulations. Overall, regions' use
of different thresholds for unacceptable underenrollment suggests that
regions may treat grantees with similar enrollment ratios differently.

ACF Annual Grantee Surveys Contain Inaccurate Enrollment Data

ACF's annual survey of grantees-the only source of nationwide information
on grantee enrollment rates-contained many inaccuracies. The PIR survey,
as it is known, requests actual enrollment figures for any 2 months that
grantees choose to report. When we attempted to verify 2001-02 PIR
enrollment data for 19 of the grantees, we found that 8 had reported
erroneously. For 6 underenrolled grantees, we found they underreported
their enrollment ratio by an average of 25 percent. We also found that 2
overenrolled grantees had erroneously reported enrollment ratios that were
over 200 percent. A similar review by ACF of 75 grantees and delegate
agencies10 that had reported particularly high or low enrollment levels
found that approximately half had erroneously reported their actual
numbers. GAO and ACF found a variety of causes that grantees cited for
misreported enrollments, including typographical errors, failure to report
children who were enrolled in the home-based or after

9While 100 percent of funded enrollment is required by Head Start
regulations, as a practical matter underenrollment can occur as a result
of programmatic fluctuations. Consequently, we asked the regional offices
to identify the threshold, if any, they used for determining when a
grantee's level of enrollment compared with its total funded enrollment
was "unacceptable." We then asked regional officials to report back to us
the grantees they were aware of that fell beneath this threshold.

10A delegate agency means a public or private nonprofit organization or
agency to which a Head Start grantee has delegated all or part of its
responsibility for operating a Head Start program. For the remainder of
the report, we will refer to grantees and delegate agencies as grantees
only.

school programs, and reporting on 2 months in which enrollment was not
their highest.

Regional Monitoring Efforts Employed Varied Criteria and Lacked Timely
Data

We found that ACF regional offices employed different criteria and used a
variety of data sources and approaches to determine if a grantee is
underenrolled. Given that regional offices are responsible for identifying
and monitoring underenrollment, we asked regional offices to identify
their operational criterion for an unacceptable level of underenrollment.
Of 11 regional offices we surveyed, we found that 3 did not utilize a
specific threshold to distinguish between acceptable and unacceptable
underenrollment, while the other 8 offices used different thresholds. Each
of the 3 regions that did not have a set threshold for "unacceptable"
underenrollment indicated that underenrollment was treated on a
case-bycase basis that would take into consideration the degree of
underenrollment and other factors, including the grantee's efforts to
increase enrollment. For the regions that specified thresholds of
"unacceptable" underenrollment, these thresholds ranged from any
enrollment ratio below 100 percent in 3 regions to below 74 percent in one
region (See table 1.)

Table 1: ACF Regional Thresholds for Unacceptable Levels of
Underenrollment

Thresholds for unacceptable ACF region or branch and coverage area
underenrollment

Region I No threshold

Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, and
Vermont

Region II No threshold

New Jersey, New York, Puerto Rico, and U.S. Virgin Islands

Region III Less than 97 percent

Delaware, District of Columbia, Maryland, Pennsylvania, Virginia, and West
Virginia

Region IV Less than state averagea

Alabama, Florida, Georgia, Kentucky, Mississippi, North Carolina, South
Carolina, and Tennessee

Region V Less than 100 percent

Illinois, Indiana, Michigan, Minnesota, Ohio, and Wisconsin

Region VI No threshold

Arkansas, Louisiana, New Mexico, Oklahoma, and Texas

Region VII Less than 74 percent Iowa, Kansas, Missouri, and Nebraska

Region VIII Less than 95 percent

Colorado, Montana, North Dakota, South Dakota, Utah, and Wyoming

Region IX Less than 95 percent

Arizona, California, Hawaii, Nevada, and Pacific Insular Areas

Region X Less than 100 percent Alaska, Idaho, Oregon, and Washington

American Indian-Alaska Native Program Branch- Less than 100 percent 23 of
50 states

Source: GAO survey of ACF regional offices and ACF.

aState average in region IV refers to the average enrollment level of all
Head Start grantees for each state within the region.

ACF regional offices reported that they identify unacceptable
underenrollment primarily by visiting grantees every 3 years and also by
engaging grantees in periodic dialogue. More than half of the regions also
said that they relied heavily on PIR data and on their review of
grantrefunding applications. Finally, 5 regions indicated that they rely
to a great

extent on their reviews of annual audits of grantees. Table 2 presents the
extent to which regional offices rely on various approaches to identify
underenrollment.

Table 2: Number of ACF Regions Relying on Various Methods to Oversee Head
Start Grantee and Delegate Agency Enrollment Levels

                               Extent relied upon

                                           Great or             Some, little, 
                                          very great  Moderate          or no 
                       On-site monitoring          11         0 
          Periodic dialogue with grantees          10         1 
               Review of grant re-funding                       
                              application           8         2 
                         Analyze PIR data           7         3 
                  Review of annual audits           5         3 

Source: GAO survey of ACF regional offices.

Each of the approaches used by regional offices to monitor enrollment is
lacking in timeliness or accuracy, or is not used systematically to
monitor underenrollment. For example, in 3 regions we visited, ACF
officials commented that while the on-site visits are designed to
systematically assess underenrollment, the visits do not provide timely
information because they are only conducted every 3 years. Conversely,
while most regional officials we surveyed said that they rely on periodic
discussions with grantees to identify underenrolled grantees, regional
officials we visited said that they do not systematically discuss
enrollment levels with grantees during this process. Officials from one
region we interviewed also said that enrollment data included in grant
re-funding applications are not informative because the data are based on
forecasts. Also, while surveyed officials listed the PIR data as a key
resource, those we spoke with said it was not necessarily accurate or
timely due to the fact that data arrive after the subsequent program year
has begun. Finally, regarding the use of annual audits, regional officials
we spoke with said they did not always receive them for all grantees and
that the audits they did receive do not necessarily comment on grantee
enrollments.11

11Currently, under the Single Audit Act any state or local government or
nonprofit organization that spends $300,000 or more per year ($500,000 for
fiscal years ending after December 31, 2003) in federal funds must have
its financial statements, internal controls, and compliance with federal
laws and regulations audited.

National Data on the Extent of Underenrollment Differs from Regionally
Reported Data on Number of Unacceptably Underenrolled Grantees

Applying a range of underenrollment thresholds to national data indicates
that a higher percentage of grantees may be underenrolled than what was
reported to us by regional offices. Regional offices, using a range of
enrollment thresholds, reported to us that about 7 percent of grantees
were unacceptably underenrolled. Comparatively, PIR survey data indicated
that more than 50 percent of Head Start grantees had enrollment ratios
below 100 percent-the regulatory definition of fully enrolled. PIR data
also showed a significantly higher proportion of grantees-33
percent-reported enrollment ratios below 95 percent than the 7 percent of
grantees reported as unacceptably underenrolled by the regional offices.
Finally, PIR data showed that a similar proportion of grantees- about 9
percent-reported an enrollment ratio below 80 percent as the 7 percent of
unacceptably underenrolled grantees reported to us by the regions. (See
fig. 5.)

Figure 5: Comparison of Enrollment Ratios as reported in 2001-02 PIR
Survey with Unacceptably Underenrolled Grantees as Reported by ACF Regions

Percentage of all grantees

                                    60 57 50

                                       40

                                       30

                                       20

            10 0 PIR < 100% PIR < 95% PIR < 80% Regionallyidentified

Source: GAO analysis of ACF regional office survey responses and PIR data.

The portion of grantees that regions reported as unacceptably
underenrolled differed from what would have been identified by applying
the regional threshold to national PIR data. When compared with the
percentage of unacceptably underenrolled grantees reported by regions,

PIR data show larger percentages of grantees below these thresholds for
each region that specified a threshold. This was true even in regions that
indicated they relied on PIR data to a great or very great extent. (See
table 3). For example, region V, using a threshold of 100 percent for
unacceptable underenrollment, reported to us that slightly less than 2
percent of its grantees were unacceptably underenrolled in 2001-02. PIR
data from that same year indicate that about 62 percent of region V
grantees had enrollment ratios less than 100 percent-a difference of 60
percentage points from what was reported to us. In fact, only regions III
and VII, of the 7 regions in table 3 with clearly defined thresholds for
unacceptable underenrollment, reported to us a percentage of unacceptably
underenrolled grantees that was within 10 percentage points of what PIR
data show using the same threshold. While we do not think that PIR data
are reliable for reporting national enrollment figures, the regional
offices based what they reported to us in part on their review of PIR
data.

Table 3: Percentage of Head Start Grantees, by Region, Reported as
Unacceptably Underenrolled by ACF Regions with Unacceptable
Underenrollment Thresholds, Compared with PIR Data at the Same Thresholds

                                                                Percentage of 
                                                                     grantees 
                                                                    reporting 
                                                                   enrollment 
                                   Percentage of               ratios at less 
                                                                         than 
                                        grantees           regionally defined 
                                        reported           
                  Regionally      by ACF regions Extent to       unacceptable 
                  defined                     as   which   
                   unacceptable   being          regional     underenrollment 
                                  unacceptably    office   
                                   underenrolled reported     threshold using 
                  underenrollment             in relying              2001-02 
                                                 on        
Region/program    threshold           2001-02 PIR data                     
       branch                                                      PIR survey

              Region V              Less than 100 percent      1.7  Moderate  
    Illinois, Indiana, Michigan,                                   
        Minnesota, Ohio, and                                       
              Wisconsin                                            

Region X Less than 100 percent 13.6 Great

Alaska, Idaho, Oregon, and Washington

American Indian-Alaska Less than 100 percent 44.5 Great
Native Branch - 23 of 50
states

             Region III            Less than 97 percent     14.3  Very great  
       Delaware, District of                                     
        Columbia, Maryland,                                      
    Pennsylvania, Virginia, and                                  
           West Virginia                                         

            Region VIII            Less than 95 percent         4.3   Some    
     Colorado, Montana, North                                       
       Dakota, South Dakota,                                        
         Utah, and Wyoming                                          

              Region IX             Less than 95 percent      0.9  Moderate   
    Arizona, California, Hawaii,                                  
     Nevada, and Pacific Insular                                  
                Areas                                             

Region VII Less than 74 percent 2.2 Great

Iowa, Kansas, Missouri, and Nebraska

Source: GAO analysis of survey responses from ACF regional offices and PIR data.

Differing Definitions of As a result of differences in regional
definitions of what constitutes an Underenrollment Create unacceptable
level of underenrollment, grantees with similar levels of Potential for
Uneven underenrollment may be treated differently across regions,
particularly in Treatment of Grantees areas without a defined threshold.
Regional offices reported to us that

they take a variety of actions to address unacceptable
underenrollment,across Regions including increased monitoring, technical
assistance, and, occasionally, enforcement actions, including recouping
funds and reducing future grant

awards. To the extent that differing thresholds affect the identification
of deficiencies that would lead to these actions, regions may subject
grantees to different treatment. For example, as shown in table 4,
although a higher percentage of grantees in region II have enrollment
levels below 95 percent than in region III, according to PIR data (44
percent versus 18 percent), region II considers only 3 percent of its
grantees unacceptably underenrolled, while region III considers 14 percent
of its grantees unacceptably underenrolled. This discrepancy may be
attributable to the fact that region II lacks a threshold for defining
unacceptable enrollment, while region III has set a threshold of 97
percent. As a result, more grantees in region III have been subject to
monitoring and enforcement actions.

 Table 4: Comparison of Region II and III Grantees Identified as Underenrolled

                                             Region III-Delaware, District of
            Region II-New Jersey, New York, Columbia, Maryland, Pennsylvania,
             Puerto Rico, and U.S. Virgin Islands Virginia, and West Virginia

Percentage of grantees with enrollment ratio                
            below 95% according to PIR              44.4 %             17.6 % 
       Threshold for unacceptable enrollment     No threshold   Less than 97% 
     Percentage of grantees reported to us by                  
       region as unacceptably underenrolled           2.7      

Source: GAO analysis of survey responses from ACF regional offices and PIR
data.

Regional and Grantee 	ACF regional officials and officials of
underenrolled Head Start grantees often cited a mixture of factors that
made it difficult to achieve full

  Officials Often Cited Combinations of Factors as Responsible for
  Underenrollment

enrollment, including increased parental demand for full-day child care
and a decrease in the number of eligible children. Many said welfare
reform has increased the number of working parents, increasing demand for
full-day child care and reducing the number of eligible children. Also,
more than one-half of the grantees we interviewed reported they were
having difficulty acquiring and developing adequate facilities. Meanwhile,
underenrolled grantees and ACF regional officials also said that
underenrollment was occurring because more parents were seeking services
with other early education and child care programs, some of which
subsidized care with relatives. Other contributing factors, such as
eligible families moving from the service area, language, and cultural
differences between children's families and program staff, and weak or
inadequate recruiting efforts by the grantees, were less frequently cited.

Multiple Factors Often Linked to Underenrollment

Many grantees indicated that the combination of multiple factors had
fostered underenrollment for their program. Nearly two-thirds of the
underenrolled grantees we spoke with cited two or more contributing
factors. For example, one northern California grantee believed that
underenrollment was caused by a decrease in income-eligible children in
its area, because the high cost of living and a shortage of affordable
housing in the area, and also by the number of families moving from
welfare to work. In addition to citing the decrease in eligible children,
this grantee expressed a need for more full-day slots, and reported facing
increasing competition from day care programs that reimbursed relatives or
friends to provide full-time child care. Similarly, one New Jersey grantee
experiencing problems acquiring a new facility was also affected by a
state supreme court decision requiring free preschool for poor children.
Additionally, this grantee felt that it was losing eligible children as a
result of families on welfare finding jobs and needing more full-day
slots. A commonly cited combination of factors-cited by 8 of the 25
grantees we interviewed-was the simultaneous shortage of full-day slots
and the movement of families out of welfare and into the workforce.

Increased Demand for Full-day Care, Facilities Problems, and Increased
Availability of Other Programs Most Frequently Cited as Affecting
Underenrollment

Grantees and Regions Said Movement of Families from Welfare to Work
Affected Enrollment and Increased Demand for Full-day Care

Regional and grantee officials most frequently cited the increased demand
for full-day child care, construction delays and inadequate facilities,
and the increased availability of early education and child care programs
as the factors causing underenrollment. Other factors, such as high
turnover rates and income eligibility criteria were also cited, but less
frequently. Each of the factors affecting underenrollment that grantees
and regions cited is described in more detail in the following sections.
Appendix II lists the factors identified by regions as contributing to
grantee underenrollment, and appendix III lists factors identified by
grantees.

Both regional and grantee officials said that the movement of low-income
families from welfare to work had contributed to underenrollment. Seven of
11 regions cited the movement of low-income families from welfare to work
as either a major or a moderate reason for grantees' underenrollments.
Similarly, of the 25 grantees we contacted, 11 cited this factor. Regional
officials and grantees suggested the movement from welfare to work
affected enrollments in two ways. First, as many parents began to work
full-time, they increasingly needed full-day care. When Head Start
grantees could not meet this need, some eligible families secured child
care elsewhere. Second, some families entered work and earned income that
disqualified their children from Head Start programs. A number of grantees
related specific examples of how the movement from welfare to work
affected enrollments. For example,

Construction Delays and Inadequate Facilities Affected Enrollments

Other Early Education and Child Care Programs Can Affect Enrollments

o  	A large grantee in Illinois said that many former welfare recipients
who need full-day child care services no longer qualify for Head Start
because they earn wages just above the Head Start income guidelines or
work rotating schedules to avoid using formal child care services.

o  	A grantee in California said that the cost associated with switching
to fullday care sometimes is a barrier to meeting families' needs.

Of the 25 underenrolled grantees we surveyed, 14 reported that difficulty
acquiring and developing adequate facilities contributed to
underenrollment. Similarly, over half of the 11 regions reported that
underenrollment was linked to a major or moderate extent to facilities
being completed more slowly than expected. For example, an Eastern grantee
was unable to serve children in need of full-day care because it lacked
classrooms and found it difficult to acquire more space. In the Midwest, 2
grantees reported that their inadequate facilities kept them from filling
about 1,800 funded slots-43 percent of their funded slots- even though
many eligible families desired Head Start services for their children. The
grantees said that they had difficulty acquiring alternate facilities:
some potential sites were environmentally unsuitable, while others faced
neighborhood opposition. In another case, an American Indian grantee that
had 25 unfilled Early Head Start slots expects to achieve 100 percent
enrollment in the fall of 2003 when a new facility is scheduled to open.

Regional and grantee officials often indicated that competition from other
early education or child care centers serving low-income preschool
children contributed to Head Start underenrollment. Seven of 11 regions
cited this factor as a major or moderate contributor to underenrollment,
and 8 of 25 grantees we interviewed identified this factor. In addition, 5
grantees said that a closely related factor also reduced families' use of
Head Start-the availability of state subsidies to pay relatives or friends
for child care.

Officials in 2 regions provided specific examples of increased
availability of other programs having a negative impact on Head Start
enrollment. According to region V officials, the availability of other
programs had a major impact on a large grantee in Michigan when the public
school system increased its preschool programming and as a result
increased the options available to Head Start-eligible children. As a
result, the grantee sustained a shortfall of almost 2,000 children.

Grantees also reported specific examples of increased availability of
other programs having a negative impact on Head Start enrollment:

o  	A large underenrolled grantee on the East Coast said that availability
of prekindergarten programs at public and charter schools is the most
important reason its delegate agencies are underenrolled.

o  	A medium-sized grantee in Oklahoma with 454 funded slots indicated
that in the 2001-02 school year, the local public school started a
preschool program for 4-year-old children that resulted in a slight
decline in Head Start enrollments at some of its service centers.

o  Officials representing a smaller grantee in Georgia with 161 funded
slots

said that their program was affected in 2001-02 when the state funded a
prekindergarten program in public schools. Specifically, the grantee said
that some children who had been pre-enrolled for Head Start switched to
the state-funded program.

Regional and grantee officials also indicated that state subsidies for
unlicensed child care caused some grantees to be underenrolled. Officials
of region IX (Arizona, California, Hawaii, Nevada, and Pacific Insular
Areas) said that increasingly, low-income parents make use of state child
care subsidies to pay for child-care exempt from licensing standards, such
as care provided by friends or nonresident relatives. Region IX officials
believed this had a significant impact on reducing Head Start program
enrollments. In another example, a grantee in Pennsylvania saw its
enrollments drop after the state allowed parents to use state child care
subsidies to pay nonlicensed child care providers such as relatives and
friends. In another instance, a grantee in California said that since its
program primarily offers part-day/part-year services, many families chose
to use subsidized, license-exempt care by relatives or friends who can
provide full-day or part-day care.

Less frequently, regions and grantees also cited other factors as
negatively influencing Head Start enrollment. Eight grantees indicated
that eligible families had moved from their service areas, often because
of the high cost of living, increasing underenrollment. Three regions also
reported that high turnover rates among enrolled children contributed
moderately to underenrollment.

Five grantees indicated that the income-eligibility criterion for Head
Start was too low in their high-cost areas. For example, 4 underenrolled
California grantees said that even relatively poor families were
disqualified from Head Start participation because their incomes, though
inadequate to meet the basic costs in the local area, were above the
federal poverty guidelines. Officials of an underenrolled grantee in
Oakland indicated that during the 2002-03 program year they had denied
Head Start services to over 100 families because their incomes exceeded
the current poverty

Other Factors Affected Enrollments, but Were Cited Less Frequently

guidelines. The other three California grantees said that they were also
turning families away because they were slightly over the income
guidelines. Grantees said that families just over the federal poverty
guideline cannot afford to send their children to education and child care
programs equivalent in quality to Head Start programs.

Four regions cited inadequate program management factors, such as weak
recruitment efforts, as a major or moderate contributor to
underenrollment. Five grantees also cited such factors as inhibiting Head
Start enrollments. Inadequate program management was characterized by weak
recruitment efforts, not developing or using waiting lists of Head
Start-eligible children, and planning enrollment expansions poorly. For
example, a grantee in Pennsylvania agreed to expand enrollment by 144
slots, and although the grantee received increased funding in the
2000-2001 program year, grantee officials said they had difficulty filling
the additional slots because of inadequate planning by the previous
management team.

Three grantees in California noted that language and cultural differences
between eligible Head Start families and program staff complicated
outreach and consequently reduced enrollments from some minority groups.
One grantee indicated that families in its service area spoke over 25
languages at home. Another grantee said it was difficult to find staff
that spoke the same languages as the families needing service.

ACF national and regional offices and grantees all report taking action to
address underenrollment, such as issuing guidance, increasing monitoring,
and attempting to conduct broader outreach efforts. The ACF national
office issued a memorandum instructing regional offices to address
underenrollment, and all ACF regions we surveyed said that they have
increased their monitoring efforts. Some ACF regions have also taken
action to reduce grantees' funding and recoup federal funds. Many grantees
we spoke with have increased outreach efforts, sought partners to help
provide more full-day services, and increased the capacity of physical
facilities. While 18 of the 25 grantees we contacted had made progress
toward achieving full enrollment, others cited continuing challenges.

  ACF and Grantees Use a Variety of Approaches to Address Underenrollment

ACF-Issued Guidance for Managing Underenrollment Lacks Specific Criteria
for Priority Review and Corrective Action

In April 2003, ACF headquarters issued policy guidance to its regional
offices instructing them to take specific actions with underenrolled
grantees, although it provided no particular instructions for the review
process or any criteria for prioritizing grantees for corrective action.
The guidance instructs regional officials to address underenrollment
depending on four possible causes. For example, if the grantee can
demonstrate that an inappropriate program option is causing
underenrollment, the guidance instructs regions to carefully consider
grantee requests to make changes to their services, such as converting
current part-day slots to full-day slots. The four causes identified in
the guidance and the recommended actions are summarized in table 5.

        Table 5: April 2003 National Head Start Guidance to ACF Regions

Category of underenrollment Expected regional action

Temporary in nature (e.g., awaiting completion Ask grantee to document in
writing when it will return to full enrollment. Periodic

of a new facility).

follow-up by assigned regional program specialist. Reevaluate grantee
status if it remains in this category for more than several months.

Attributable to a nonimplemented expansion Generally, all grantees must
have the additional children enrolled in their
(i.e., where a grantee that was given programs within 1 year of receiving
their grant expansion. Regions should
expansion funding to serve an increased  o  track grantee expansion,
number of children failed to enroll these

children within a reasonable time period).  o  contact grantee and discuss
reasons for delays,

o  judge whether grantee is making sufficient progress to warrant
extension,

o  	require an implementation plan for any extension, which must not
exceed 6 months, and

o  	inform the grantee in writing that expansion funds will no longer be
available if at any time the region determines that the grantee will not
be able to implement its approved expansion in a reasonable time period.

Attributable to demographic changes that have Determine an appropriate
reduction in the grantee's enrollment and funding reduced the number of
eligible children in the levels. ACF's general policy will be to

grantee's service area.  o  give the grantee appropriate notice that a
funding reduction will be initiated,

o  	implement funding reductions at the time a grant is being refunded,
based on a grantee's historical underenrollment problems, and

o  	reduce funding proportionate to the degree of underenrollment (i.e.,
on a cost per child basis adjusting for those grantee costs that are not
directly related to enrollment such as the salaries and fringe benefits of
management staff.

Not attributable to any of the above causes, but occurring because of
grantee management issues. This would include poor community outreach,
inadequate needs assessments, inappropriate program options, inadequate
transportation services, poor facility planning, lack of coordination with
other community providers, such as prekindergarten programs, or any other
management problems causing underenrollment.

Make an on-site monitoring visit to the grantee to fully assess the
reasons for underenrollment. The assessment could result in the region
designating the grantee as deficient. (The region may designate a grantee
as being in noncompliance rather than deficient if it determines
underenrollment is an isolated issue that does not seem to be part of a
more systemic problem with the grantee's ability to provide an appropriate
level of Head Start services.)

A deficient grantee is required to submit a Quality Improvement Plan (QIP)
indicating how, within 1 year or less, it will achieve full enrollment.
This could be accomplished by fixing the problem causing the
underenrollment or by agreeing to an enrollment reduction that would bring
the grantee to its full enrollment level, or some combination thereof. The
region should monitor the grantee's progress in implementing its QIP and
provide any appropriate technical assistance. If the underenrollment has
not been corrected at the end of the QIP period, the region needs to
initiate an adverse action against the grantee, which would be termination
or denial of re-funding.

If an inappropriate program option is a major factor causing
underenrollment, regions must consider a grantee's request to reconfigure
its program options. Changes such as conversion of part-day slots to
full-day or reduction of double sessions may include some proposed
reduction for enrollment levels. Each such proposal should be judged on
its own merits, including the extent to which the grantee proposes to
collaborate with other community providers and the extent to which the
proposed reconfiguration is supported by data from the grantee's current
community assessment.

Source: GAO analysis of ACF memorandum.

The April guidance does not suggest any systematic process for identifying
underenrolled grantees, nor does it specify criteria for prioritizing when
grantees should be subject to corrective action based on their level of
underenrollment. One regional official said that the lack of a threshold
offered no gauge for establishing priorities and intensifying monitoring
efforts.

Regional Officials Reported Intervening with Underenrolled Grantees to
Correct Underenrollment

The ACF regions we surveyed reported taking a variety of actions to
address underenrollment ranging from providing assistance to recouping
federal funds in some cases. Officials in all 11 regions responded that
they had taken at least one action to ensure that grantees address
underenrollment. The interventions taken most often were to monitor
enrollment levels (55 grantees), track improvement efforts (43), and
provide training and technical assistance (30). Notably, 4 regions
provided additional funds to a total of 18 underenrolled grantees to
purchase or renovate facilities. Somewhat less often, regions took action
to reduce funded enrollment levels or recoup funds. Specifically, only 2
regions reported that they recouped funds from a total of 6 underenrolled
grantees. (See table 6.)

         Table 6: Actions by ACF Regions toward Underenrolled Grantees

                                         Count of regions  Number of grantees 
                                                                           to 
             Regional actions            taking the action       which action 
                                                                      applied 
Identified deficiency and pursued QIP                 6 
Met with grantees and developed plan                  5 
        to address underenrollment                         
    Required frequent (such as monthly)                  5 
          reporting of enrollment                          
      Provided funds for purchase or                       
renovation of facilities or for more                    
               comprehensive                               
           community assessments                         4 
      Tracked correction efforts for                     3 
                compliance                                 
      Provided training and technical                    3 
                assistance                                 
     Negotiated reductions in funding                    3                  9 
          levels and funded slots                          
Withheld or recaptured funds for the                  2                  6 
          year of underenrollment                          
    Held meetings with grantee's board                   2                  3 
              and management                               
       Proposed Head Start grant be                      2                  2 
        relinquished or terminated                         
Requested audit by the HHS Office of                  1                  1 
             Inspector General                             
          Changed program options                        1                  1 

Source: GAO analysis of ACF regional office survey responses.

As noted earlier, many regions and grantees said the need for full-day
services was a major factor fostering underenrollment. As part of their
efforts to assist grantees in providing more full-day services, some ACF
regional officials told us they had encouraged grantees to collaborate
with other programs or had provided additional funds to purchase or
renovate facilities. However, such efforts can be costly. For example,
region V officials told us that it costs more to provide full-day care
than part-day care because full-day care requires more facility space and
staff per child.

Underenrolled Grantees Report Taking Some Remedial Actions

Grantees we interviewed took a variety of actions to address
underenrollment, including more aggressive recruiting efforts,
collaborating with other preschool and child care programs, and increasing
slots in selected program options such as home-based services. Most
grantees we contacted said that they had taken one or more actions. The
most frequently mentioned was more aggressive recruiting followed by
collaboration with other programs. For example, a large grantee in New
York State that faced increased demand for full-day care since welfare
reform collaborated increasingly with other child care providers to piece
together a package of full-day services. Nine grantees also reported
trying to increase physical facilities capacity. (See table 7.) Other
actions taken to address underenrollment, which were cited by 3 or fewer
grantees, included improving the tracking or monitoring of enrollment
opportunities, hiring multilingual staff, reducing the number of funded
slots, and providing contractual incentives for delegate agencies to
maintain full enrollment.

Table 7: Actions Taken by Interviewed Grantees to Address Underenrollment

Number of grantees taking the action

Action taken (n = 25)

Took more aggressive and proactive recruiting approach

Collaborated with other preschool and child care programs

Increased capacity of physical facilities

Increased slots in selected program options

Worked on updating community assessment

Increased marketing of Head Start to the community

Relocated and developed new program centers

Identified a new unserved low-income population

Trained staff in recruitment and program promotion

Increased home-based program enrollments

        Source: GAO analysis of interviews with underenrolled grantees.

Grantees Cite Obstacles to Providing Additional Full-Day Care

Several grantees told us that converting part-day services to full-day was
often challenging to implement in addition to being more costly.
Additionally, 2 grantees said that ACF did not fully understand all that
was involved in transitioning from part-day to full-day services, and that
there was no clear national guidance on how to do so. According to these
California grantees,

The costs of transitioning part-day, double sessions, to full-day services
have never been fully understood and no national process has emerged to
assist grantees and regional offices to address this problem. The major
costs often include facilities and additional staffing (where only two and
one-half staff are needed for a double session, four to six are needed to
staff a full-day session, depending on the number of hours the option
operates). Such fixed costs would require a reduction in the number of
slots (children enrolled) or an increase in funding in order to transition
from part-day double sessions to two full-day sessions.

While these two grantees expressed concern over a lack of guidance, it
should be noted that there is national guidance on budgeting for
partnerships between child care and Head Start and on financial

management issues in Head Start programs utilizing other funding
sources.12

Furthermore, grantees told us of their concern to maintain total funded
enrollment levels, even as they were converting unfilled part-day openings
to full-day. According to region V officials, this concern to maintain
enrollment levels may be in keeping with national efforts to serve a
greater number of needy children.13 For example, one underenrolled grantee
said that ACF suggested several alternatives to address underenrollment,
including converting part-day to full-day slots, but would not permit the
grantee to reduce funded slots as a way to address underenrollment.
Consequently, while converting part-day slots to full-day slots, the
grantee would have had to expand its facilities or find other child care
partners in order to serve the same number of children.

Grantees Reported Mixed Results Resolving Underenrollments

Conclusions

Some grantees reported success addressing underlying factors contributing
to underenrollment, while others did not. Of the 25 underenrolled grantees
that we contacted, 18 (72 percent) indicated that their underenrollment
had either been corrected (10 grantees) or would be corrected shortly (8
grantees). These 18 grantees overcame a variety of factors that they said
affected underenrollment. For example, 6 of these 18 grantees overcame a
shortage of available full-day slots and 8 managed to fill slots lost due
to a decline in eligible children attributed to declining TANF rolls. The
7 grantees that had not made progress addressing underenrollment often
cited similar issues. For example, 3 of these 7 grantees said they had
faced challenges resulting from decreasing TANF caseloads and were unable
to respond to the increased demand for fullday services. On the basis of
our limited number of interviews, we could not determine why some grantees
reported they were able to successfully address problems that other
grantees could not.

Because ACF has no reliable nationwide data on enrollment, it is not
possible for the agency to identify and track underenrollment trends and
to develop strategies to ensure that federally funded Head Start slots are
filled. While we could not determine with any precision the extent to

12ACF Information Memorandums: IM-HS-01-13, 11/16/2001 and IM-HS-01-06,
3/8/2001.

13ACF's primary performance indicator for the number of children served
gives equal weight to part-day and full-day slots.

which there is underenrollment, our survey work and analysis indicate it
is possible that underenrollment is more widespread than ACF has
acknowledged. The complexity of factors buffeting Head Start grantees
underscores the need for ACF to accurately identify underenrollment and
its causes on a timely basis. Even if ACF corrects national survey data
issues, there is no guarantee that its regions will know of
underenrollment in a timely manner because the main national data source
is not available until the following program year.

Furthermore, because ACF regions vary in how they define unacceptable
levels of underenrollment and because they rely on approaches to identify
grantees that are not timely or consistent, there is some indication that
Head Start grantees with similar levels of underenrollment are treated
differently across regions. ACF guidance to the regions on how to address
different types of underenrollment is a good first step toward a more
systematic approach to underenrollment. However, until ACF issues guidance
that more clearly explains how to prioritize grantees with varying levels
of underenrollment for purposes of corrective action, regions are likely
to continue using varied criteria or none at all. Also, until more timely
and systematic approaches are developed for regions to identify
underenrolled grantees, it is possible that low enrollment will go
undetected and federal dollars will not be fully utilized for low-income
children who could benefit from Head Start's program goals.

Finally, it appears that there may be a perceived incentive for
underenrolled grantees to maintain or increase enrollments due, in part,
to ACF's emphasis on counting the total number of children served
irrespective of whether they are enrolled part-day or full-day. Measuring
Head Start enrollments without capturing the difference in level of
service provided by full-day or part-day programs adds to the difficulty
of meeting local needs and adjusting to changes in those needs. Until ACF
can get a better grasp of the nature and size of underenrollment and align
program incentives with family needs, it may be a challenge for Head Start
to best meet the needs of some families it could serve.

Recommendations for 	We recommend that the Secretary of HHS direct ACF to
(1) take steps to ensure the accuracy of enrollment data reported in its
annual nationwide

Executive Action 	survey of grantees, (2) develop a standard criterion for
regional offices to use in identifying grantees whose underenrollment
merits monitoring or corrective actions, (3) develop an additional measure
of aggregate services other than total enrollment that takes into
consideration the different levels of service provided by full-day and
part-day programs, and (4) work

with regional offices to develop a more systematic process for them to
collect reliable enrollment data during the program year so that they can
address underenrollment more quickly.

Agency Comments 	We provided a draft of this report to the Department of
Health and Human Services for review and comment. In its written response,
included as appendix IV of this report, HHS agreed with our
recommendations and indicated that it will take action to address each
recommendation.

We are sending copies of this report to the Secretary of Health and Human
Services and other interested parties. In addition, the report will be
available at no charge on GAO's Web site at http://www.gao.gov.

If you or your staffs have any further questions about this report, please
call me on (202) 512-7215. Other GAO contacts and staff acknowledgments
are listed in appendix V.

Sincerely yours,

Marnie S. Shaul Director, Education, Workforce, and Income Security Issues

                       Appendix I: Scope And Methodology

To determine what is known about the extent to which Head Start programs
are underenrolled, we assessed the reliability of PIR enrollment data,
conducted interviews, reviewed program documentation, and surveyed ACF
regional offices. Specifically, we assessed the reliability of the PIR
data on grantee enrollment by (1) performing electronic testing of key
data elements for obvious errors in completeness and accuracy, (2)
reviewing existing information about the data and the system that produces
it, and (3) contacting 19 underenrolled and overenrolled grantees selected
across a range of reported enrollment ratios. We did not assess the
reliability of other PIR data used in the report. We also interviewed ACF
headquarters officials, reviewed federal guidance and regulations on
enrollment, surveyed the regions and a branch office, and interviewed
regional officials in regions III, V, and IX. The 3 regions were selected
for site visits on the basis of geographical representation and the number
of underenrolled grantees they reported to us. Since ACF oversight of Head
Start grantees is primarily accomplished through its regions and program
branches, we designed a survey instrument in which these entities could
provide written responses to our specific requests for such information
as:

o  	the methods the regions used to oversee grantee and delegate agency
enrollment levels;

o  	the threshold, if any, they had established for determining the point
at which a grantee or delegate agency's level of underenrollment is
considered to be unacceptable;

o  	a list of all grantees and delegate agencies that they believed had
unacceptable levels of underenrollment for both the 2001-02 and 2002-03
program years;

o  	the reasons that they believed unacceptable levels of underenrollment
had occurred and the extent (major, moderate, minor, or none) that they
believed each identified reason had contributed to underenrollment;

o  	the actions they had taken to address the unacceptable level of
underenrollment for their grantees and delegate agencies.

We surveyed all 10 ACF regional offices and the American Indian-Alaska
Native Program Branch. The Migrant and Seasonal Program Branch was
excluded from our review because of its lack of comparability with the
other branch and regions caused by anticipated enrollment fluctuations
resulting from the seasonal movement of migrant families.

Appendix I: Scope And Methodology

To determine ACF officials' and Head Start grantees' views on the factors
that contribute to underenrollment and to identify actions they took to
address underenrollment, we relied on our survey of the ACF regional
offices and the American Indian-Alaska Native Program Branch office as
well as a interviews with 27 grantees (1 of which was actually a delegate
agency) that had been identified by the regions as underenrolled. Twenty
grantees were contacted by telephone and 7 were interviewed face-to-face.
Two of the 27 grantees said that they had not experienced any
underenrollment; therefore, our grantee survey results were based on the
responses of 25 grantees that agreed with the regions' designation of
their underenrolled status.

Using a standard set of questions, we interviewed at least 1 identified
grantee from each region or branch. In selecting grantees to be
interviewed, we chose 7 from the metropolitan areas of the 3 regions that
we visited so that we could conduct some of the interviews in person. The
other 20 grantees we interviewed were primarily selected from each region
based on having been identified as being underenrolled for 2 program years
(2001-02 and 2002-03). We also attempted to interview both grantees that
were funded for more than 500 slots and grantees that were funded for
fewer.

The grantee interview requested that grantee officials

o  	describe the factors they believed contributed to the grantee's
underenrollment,

o  identify the actions the grantee had taken to address underenrollment,
and

o  	indicate whether they believed that the grantee's underenrollment had
been corrected.

Because of the lack of reliable enrollment data, the information we
collected regarding underenrollment was primarily testimonial. Apart from
assessing the basic consistency of interviewees' responses with known
program characteristics, we did not independently test the information
they provided, such as reasons for underenrollment.

Appendix II: Factors That the ACF Regions Believed Contributed to
Underenrollment to a Major or Moderate Extent

Table 8: Factors Cited By ACF Regions as Contributing to Underenrollment
to a Major or Moderate Extent, during Program Years 2001-02 and 2002-03

                                                                Percentage of 
                                               Number of times regions citing 
         Factor cited as contributing to                 cited         factor 
                 underenrollment                               
    Fewer eligible families (moving into jobs                7 
                  and off TANF)                                
            Not enough full-day slots                        7 
     Other day care or education centers are                 7 
              available in the area                            
    Completion of facilities was slower than                 6 
                     planned                                   
Demographic change: decrease in the number                6 
      of eligible children in service area                     
     Shortage of eligible children below the                 5 
               poverty rate cutoff                             
             Too many part-day slots                         5 
     Waiting lists of eligible children not                  4 
                developed or used                              
            Weak recruitment efforts                         4 
     High turnover rate (e.g., families move                 3 
                     often)                                    
    Actual enrollments not yet caught up with                2 
         recent funded program expansion                       
            Not enough part-day slots                        1 
             Too many full-day slots                         1 

         Source: GAO analysis of ACF regional office survey responses.

Appendix III: Factors that Grantees Believed Contributed to Their Head Start
Programs' Underenrollment

Table 9: Factors Cited by 25 Grantees as Contributing to Head Start
Program Underenrollment, during Program Years 2001-02 and 2002-03

                                                                Percentage of 
                                                              grantees citing 
                                             Number of times           factor 
Factors contributing to underenrollment             cited         (n = 25) 

Difficulties in acquiring and developing adequate facilities 14

Families moving into jobs and off TANF 11

Not enough full-day slots 9

Other day care or education centers are in the area 8

Lack of housing for low-income families 7

Income eligibility criterion too low for high-cost area 5

Programs paying relatives or friends for child care 5

Lack of income-eligible children in service area 4

High turnover rate (e.g., families move often) 4

High cost of living causes low-income families to move 4

Language and cultural barriers among eligible families make selling
program difficult 3

Actual enrollments not yet caught up with expansion 2

Centers in wrong service area location 2

Centers with wrong mix of program service options 2

High cost of transitioning from part-day to full-day care 2

Inadequate program management 2

                   Poorly planned enrollment expansion                  2 8.0 
                  Weak or inadequate recruiting efforts                 2 8.0 
             Completion of facilities was slower than planned           1 4.0 
      Difficulties in finding qualified collaborative care partners     1 4.0 
Head Start income levels too low for collaborating with other        1 4.0 
low-income programs                                                    
    Parents do not recognize the benefit of Head Start for their child  1 4.0 
         Waiting lists of eligible children not developed or used       1 4.0 
     Source: GAO analysis of interviews with underenrolled grantees.      

Appendix IV: Comments from the U.S. Department of Health and Human Services

Appendix IV: Comments from the U.S. Department of Health and Human
Services

Appendix V: GAO Contacts and Staff Acknowledgments

GAO Contacts 	Betty Ward-Zukerman, (202) 512-2732, [email protected]
Bryon Gordon, (202) 512-9207, [email protected]

Acknowledgments 	In addition to those named above, Daniel Jacobsen, Lesley
Woodburn, Luann Moy, Barbara Johnson, James Rebbe, Susan Bernstein, and
Amy Buck made key contributions to the report.

Related GAO Products

Education and Care: Head Start Key Among Array of Early Childhood
Programs, but National Research on Effectiveness Not Completed.
GAO03-840T. Washington, D.C.: July 22, 2003.

Head Start and Even Start: Greater Collaboration Needed on Measures of
Adult Education and Literacy. GAO-02-348. Washington, D.C.: March 29,
2002.

Early Childhood Programs: The Use of Impact Evaluations to Assess Program
Effects. GAO-01-542. Washington, D.C.: April 16, 2001.

Title I Preschool Education: More Children Served, but Gauging Effect on
School Readiness Difficult. GAO/HEHS-00-171. Washington, D.C.: September
20, 2000.

Early Education and Care: Overlap Indicates Need to Assess Crosscutting
Programs. GAO/HEHS-00-78. Washington, D.C.: April 28, 2000.

Early Childhood Programs: Characteristics Affect the Availability of
School Readiness Information. GAO/HEHS-00-38. Washington, D.C.: February
28, 2000.

Education and Care: Early Childhood Programs and Services for Low-Income
Families. GAO/HEHS-00-11. Washington, D.C.: November 15, 1999.

Head Start: Challenges in Monitoring Program Quality and Demonstrating
Results. GAO/HEHS-98-186. Washington, D.C.: June 30, 1998.

Head Start Programs: Participant Characteristics, Services, and Funding.
GAO/HEHS-98-65. Washington, D.C.: March 31, 1998.

  GAO's Mission

Obtaining Copies of GAO Reports and Testimony

The General Accounting Office, the audit, evaluation and investigative arm
of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of
the federal government for the American people. GAO examines the use of
public funds; evaluates federal programs and policies; and provides
analyses, recommendations, and other assistance to help Congress make
informed oversight, policy, and funding decisions. GAO's commitment to
good government is reflected in its core values of accountability,
integrity, and reliability.

The fastest and easiest way to obtain copies of GAO documents at no cost
is through the Internet. GAO's Web site (www.gao.gov) contains abstracts
and fulltext files of current reports and testimony and an expanding
archive of older products. The Web site features a search engine to help
you locate documents using key words and phrases. You can print these
documents in their entirety, including charts and other graphics.

Each day, GAO issues a list of newly released reports, testimony, and
correspondence. GAO posts this list, known as "Today's Reports," on its
Web site daily. The list contains links to the full-text document files.
To have GAO e-mail this list to you every afternoon, go to www.gao.gov and
select "Subscribe to e-mail alerts" under the "Order GAO Products"
heading.

Order by Mail or Phone 	The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:

U.S. General Accounting Office 441 G Street NW, Room LM Washington, D.C.
20548

To order by Phone: 	Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

Contact:

To Report Fraud, Web site: www.gao.gov/fraudnet/fraudnet.htm

  Waste, and Abuse in E-mail: [email protected]

Federal Programs Automated answering system: (800) 424-5454 or (202)
512-7470

Jeff Nelligan, Managing Director, [email protected] (202) 512-4800

Public Affairs 	U.S. General Accounting Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***