Early Childhood Programs: Parent Education and Income Best Predict
Participation (Letter Report, 12/28/94, GAO/HEHS-95-47).
Using data from the 1990 census, GAO analyzed factors, including family
income, parents' education, race, immigrant status, family type, and
others, that influence the likelihood of children taking part in
preschool programs. GAO found that children whose parents have less than
a high school diploma, whose families have low incomes, and who live in
certain states are the least likely to participate in preschool.
Research has shown considerable benefits of preschool participation, and
yet some children remain much less likely to participate compared with
their peers. Consequently, the National Education Readiness Goal that
all children will be ready for school by the year 2000 will be difficult
to attain.
--------------------------- Indexing Terms -----------------------------
REPORTNUM: HEHS-95-47
TITLE: Early Childhood Programs: Parent Education and Income Best
Predict Participation
DATE: 12/28/94
SUBJECT: Demographic data
Children
Disadvantaged persons
Preschoolers
Head-start programs
Educational programs
Income statistics
Secondary education
Minorities
Parents
**************************************************************************
* 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. *
* *
* A printed copy of this report may be obtained from the GAO Document *
* Distribution Facility by calling (202) 512-6000, by faxing your *
* request to (301) 258-4066, or by writing to P.O. Box 6015, *
* Gaithersburg, MD 20884-6015. We are unable to accept electronic orders *
* for printed documents at this time. *
**************************************************************************
Cover
================================================================ COVER
Report to the Chairman, Subcommittee on Children, Family, Drugs and
Alcoholism, Committee on Labor and Human Resources, U.S. Senate
December 1994
EARLY CHILDHOOD PROGRAMS - PARENT
EDUCATION AND INCOME BEST PREDICT
PARTICIPATION
GAO/HEHS-95-47
Preschool Participation
Abbreviations
=============================================================== ABBREV
HHS - Department of Health and Human Services
LTHD - less than high school diploma
LI - linguistically isolated
MA - Metropolitan Area
OMB - Office of Management and Budget
WIC - Women, Infants, and Children program
Letter
=============================================================== LETTER
B-259742
December 28, 1994
The Honorable Christopher J. Dodd
Chairman, Subcommittee on Children,
Family, Drugs and Alcoholism
Committee on Labor and Human Resources
United States Senate
Dear Mr. Chairman:
In recent years, the Congress has consistently recognized the
importance of providing preschool for young children by increasing
the budget for the Head Start program. The Congress has also
recently enacted the Goals 2000 legislation. The first goal states
that by the year 2000 all children will enter school ready to learn.
One major objective of the first goal is that all children will have
access to quality preschool. Further, welfare reform proposals
discussed in the 103rd Congress included provisions for increases in
the availability of child care for low-income working women.
To assist the Subcommittee's deliberations over early childhood
programs in the context of the National Education Readiness Goal and
the ongoing welfare reform debate, you asked that we determine the
individual, family, and geographic characteristics that influence
children's preschool participation.
In July 1993, we provided you with demographic information on
preschool-aged children (3- and 4-year-olds) in our report, Poor
Preschool Aged Children: Numbers Increase but Most Not in Preschool
(GAO/HRD-93-111BR, July 21, 1993). In that report, we found that
most at-risk\1 preschool-aged children participated in preschool at
lower rates than those not at risk. More specifically, we found that
children in each of the following groups participated at lower rates
than those not at risk: children who live in immigrant families,\2
linguistically isolated\3 (LI) households, single-parent families,
families where the most educated parent has less than a high school
diploma, and families where the parents were not working. We also
found that poor children in some states participated in preschool at
higher rates than those in other states. However, that report did
not examine whether these patterns hold when simultaneously
controlling for other characteristics.
You asked us to more closely examine this information to determine
whether these patterns hold for those at risk, after controlling for
selected individual, family, and geographic characteristics. You
also asked what the relative effect of each individual, family, and
geographic characteristic is on children's preschool participation.
To answer your questions we applied a multivariate statistical
technique--a logistic regression--to data from the 1990 Decennial
Census.\4 This multivariate approach allowed us to determine the
relative effects of each of the following variables on preschool
participation: income, education status of most educated parent,
race, immigrant status, linguistic-isolation status, employment
status of parent, family type, the urbanicity of residence, and state
of residence. We did our analysis based on two logistic regression
models, one for all children and another for poor children. (See
app. I for a more detailed discussion of our objectives, scope, and
methodology and for definitions of terms used throughout the report.)
--------------------
\1 At-risk is defined as those in immigrant families, linguistically
isolated households, single-parent families, families where the most
educated parent has less than a high school diploma, or families
where the parents do not work.
\2 Immigrant child is defined as a child who is foreign born or whose
parents are foreign born and came to the United States within the
last 10 years.
\3 Linguistically isolated child is defined as a child in a household
in which no one 14 years or older speaks English "only" and no one 14
years old or older who speaks a language other than English speaks
English "very well."
\4 Our prior report provided descriptive statistics separately for
each risk group but did not control for other characteristics.
RESULTS IN BRIEF
------------------------------------------------------------ Letter :1
With Head Start funds reaching less than half the eligible 3- and
4-year-olds, we found that children living in low-income families are
less likely to attend preschool than their middle-income
counterparts.\5 In addition, the education level of children's
parents has a large influence on children's participation. After
controlling for other characteristics, children whose most educated
parent has not completed high school are less likely to go to
preschool compared with those whose most educated parent has
graduated from high school.
Surprisingly, children in some risk groups appear about as likely to
participate in preschool as those not at risk, after controlling for
other characteristics. In this regard, immigrants are about as
likely to attend preschool as nonimmigrants. Further, LI
children--those in families where no person speaks English well--are
nearly as likely to attend preschool as non-LI children, and children
whose parents are not working are nearly as likely to attend
preschool as those with working parents.
It is also interesting to note that black and American Indian
children are more likely than white children to attend preschool,
after controlling for individual, family, and geographic
characteristics. Children with single parents are also more likely
to participate than those in married-couple families, after
controlling for other characteristics.
Although it might be suggested that state variability in preschool
participation is mostly a result of the demographics of the states,
we found that demographics account for less than half of the
variability. Even after controlling for characteristics such as
income and the percentage of the children in urban versus rural
areas, children in some states are still much more likely to
participate in preschool than children in other states.
--------------------
\5 Fiscal year 1994 data indicate that approximately 40 percent of
eligible children are attending Head Start. Because our analysis was
based on 1990 decennial census data, all numbers in the report are
from 1990 unless otherwise indicated.
BACKGROUND
------------------------------------------------------------ Letter :2
In 1994, the Congress made a commitment to young children by enacting
the Goals 2000: Educate America Act. The first goal of this
legislation states that by the year 2000 all children in America will
start school ready to learn. The objectives of this goal are that:
(1) all children will have access to quality preschool, (2) all
parents will be a child's first teacher and will have access to the
training and support they need, and (3) all children will arrive at
school with healthy minds and bodies.
Research has demonstrated that children with certain risk
characteristics are significantly less likely to succeed in school.\6
Research has also concluded that children who receive high-quality
preschool services have improved test scores in elementary school,
fewer grade retentions, and reduced placements in special education
programs.\7
Furthermore, individuals who receive high-quality preschool as
children are more likely to attend college and hold jobs, and are
less likely to be on welfare or arrested for a serious crime compared
with those who do not receive such services.\8
Despite the demonstrated benefits of quality preschool for
disadvantaged children, recent studies have shown that disadvantaged
children continue to be the least likely to participate in
preschool.\9 In 1990, approximately 35 percent of poor 3- and
4-year-olds participated in preschool compared with approximately 60
percent of those in the highest income group.
Head Start is administered by the Department of Health and Human
Services (HHS) and currently targets children who live in families
below the poverty level (income below $12,674) or who receive public
assistance.\10 The program funds child development services,
coordinates nutrition and health services, and provides parental
education services. The Head Start program, funded at $3.3 billion
in fiscal year 1994, is the largest federal program providing
preschool and developmental services to poor preschool-aged children
and social services for their families. However, despite the
increasing funding in the past few years, Head Start still serves
only about 40 percent of the eligible 3- and 4-year-olds in the
country. Further, quality improvements could be made in some Head
Start centers.
In addition to Head Start, federal, state, and local governments fund
early childhood programs, some of which include parental involvement
and education as important elements. The Even Start program,
administered through the Department of Education, is one such program
designed to help parents become full partners in the education of
their children and to assist children in reaching their full
potential.
--------------------
\6 Gary Natriello, Edward McDill, and Aaron Pallas, Schooling
Disadvantaged Children: Racing Against Catastrophe (New York:
Teachers College Press, 1990).
\7 The Impact of Head Start on Children, Families and Communities,
U.S. Department of Health and Human Services, Office of Human
Development Services, Head Start Bureau, Administration for Children,
Youth and Families, Pub. No. (OHDS)-85-31193 (Washington, D.C.:
U.S. Government Printing Office, 1985).
\8 John R. Berrueta-Clement, Lawrence J. Schweinhart, W. Steven
Barnett, Ann S. Epstein, and David P. Weikart, Changed Lives: The
Effects of the Perry Preschool Program on Youths Through Age 19,
Ypsilanti, Michigan: High/Scope Educational Research Foundation,
1984; and Lasting Effects After Preschool, U.S. Department of
Health, Education and Welfare, Office of Human Develoment Services,
Administration for Children, Youth and Families, Pub. No. (OHDS)
79-30179 (Washington, D.C.: 1979).
\9 Poor Preschool-Aged Children: Numbers Increase but Most Not in
Preschool (GAO/HRD-93-111BR, July 21, 1993).
\10 Head Start uses Office of Management and Budget (OMB) poverty
guidelines. Throughout this report poverty ratios are used to
control for family size. The dollar amounts provided in this report
are the total household income for an average-sized family--a family
of four in 1990. If the family is smaller, the dollar amount is
less; similarly, if the family is larger, the dollar amount is more.
PRINCIPAL FINDINGS
------------------------------------------------------------ Letter :3
INCOME AND PARENT EDUCATION
LARGEST INFLUENCES
---------------------------------------------------------- Letter :3.1
Income and parent education are the largest influences on children's
preschool program participation. After controlling for selected
demographic characteristics,\11
children in or near poverty are 16 to 20 percent less likely to
attend preschool programs than middle-income children.\12
By contrast, children in higher-income families are 25 to 50 percent
more likely to go to preschool than middle-income children. (See
fig. 1 and table II.1.)
By contrast, a recent study shows that when publicly funded
kindergarten is available to all, children in and near poverty are
about as likely to participate as children with higher incomes.\13
This finding suggests that publicly funded kindergarten is effective
in equalizing access to programs before the first grade.
Figure 1: Poor and Near-Poor
Children Less Likely to
Participate
(See figure in printed
edition.)
Children whose most educated parent has not completed high school are
also among the least likely to attend preschool. After controlling
for selected characteristics,\14 children whose most educated parent
has less than a high school diploma are 25 percent less likely to go
to preschool than children whose parents have a high school diploma.
Compared with children whose parents have a high school diploma,
children whose most educated parent has a college degree or more are
about 46 percent more likely to attend a preschool program. (See
fig. 2 and table II.2.)
The relationship between parental education and children's likelihood
of participating in preschool may be due to a lack of knowledge about
or appreciation of the importance of early childhood education.
Figure 2: Parents with Low
Education Less Likely to Send
Children to Preschool
(See figure in printed
edition.)
Poor children whose most educated parent has not completed high
school are also less likely (23 percent) to go to preschool compared
with poor children whose parents have a high school diploma.
--------------------
\11 The characteristics we controlled were for race/ethnicity, family
type, immigrant status, linguistic-isolation status, education status
of most educated parent, work status of parents, urbanicity, and
state of residence.
\12 We define family income using OMB guidelines and Census data for
poverty increments to control for family size. The annual income
dollar amounts that correspond to the categories used in the report
for a child in a family of four in 1990 are as follows: poor: below
$12,674; near-poor 1: above $12,674 to $16,856; near-poor 2: above
$16,856 to $23,446; middle-income: above $23,446 to $44,359;
upper-middle income: above $44,359 to $63,370; highest income:
above $63,370.
\13 Access to Early Childhood Programs for Children at Risk, U.S.
Department of Education, National Center for Educational Statistics,
Office of Education , Research and Improvement (NCES 93-372)
(Washington, D.C.: 1994.)
\14 The characteristics we controlled for were race/ethnicity, family
income and type, immigrant status, linguistic-isolation status, work
status of parent, urbanicity, and state of residence.
RACE, URBANICITY, AND FAMILY
STATUS MODERATE PREDICTORS
OF PRESCHOOL PARTICIPATION
---------------------------------------------------------- Letter :3.2
BLACK AND AMERICAN INDIAN
CHILDREN ARE MORE LIKELY
TO PARTICIPATE
-------------------------------------------------------- Letter :3.2.1
Controlling for selected demographic characteristics,\15 black
children are 17 percent and American Indian children are 31 percent
more likely to attend preschool programs compared with white
children. These findings for black and American Indian children are
consistent with public policies aimed at raising the enrollment
levels of disadvantaged minority children. Although Hispanic and
Asian children are somewhat less likely to participate than white
children after controlling for other characteristics, including
poverty and education status of the most educated parent, these
differences are not material.\16 (See fig. 3 and table II.3.)
Figure 3: Likelihood of
Participating Relative to White
Children
(See figure in printed
edition.)
Controlling for other demographic characteristics,\17
differences in preschool participation for poor children are even
larger. For example, poor black children are 32 percent and American
Indian children are 60 percent more likely to go to preschool
compared with white children. Once again, these findings are
consistent with public policies aimed at raising the enrollment
levels of these children.
--------------------
\15 The characteristics we controlled for were family income and
type, immigrant status, linguistic-isolation status, education status
of the most educated parent, work status of parent, urbanicity, and
state of residence.
\16 We define material differences as those greater than 15 percent.
Because of the large sample size, virtually all differences stated in
this report are statistically significant. We therefore focus on
those differences that are material or notable.
\17 The characteristics we controlled for were family income and
type, immigrant status, linguistic-isolation status, education status
of the most educated parent, work status of parent, urbanicity, and
state of residence.
RURAL CHILDREN AMONG
LEAST LIKELY TO
PARTICIPATE
-------------------------------------------------------- Letter :3.2.2
Compared with children in rural areas and accounting for other
characteristics,\18 children in small towns are 17 percent more
likely to go to preschool, and children in smaller cities and suburbs
and the largest cities are 29 and 35 percent, respectively, more
likely to attend preschool programs.\19 (See fig. 4 and table II.4.)
Figure 4: Likelihood of
Participating Relative to
Children in Rural Areas
(See figure in printed
edition.)
Among the poor, children in rural areas are also less likely to
participate in preschool after accounting for other demographic
characteristics. However, poor children in nonrural areas are about
20 percent more likely to participate in preschool than those in
rural areas, regardless of size of town.
Research suggests a number of reasons that may explain low
participation among rural children.\20 Reasons include greater
reluctance to receive public services, greater distances to travel to
receive services, and poor public transportation.
--------------------
\18 The characteristics we controlled for were race/ethnicity, family
income and type, immigrant status, linguistic-isolation status, work
status of parent, education status of the most educated parent, and
state of residence.
\19 We define largest cities as the urban portions of the counties
comprising the 25 largest cities in 1990, smaller cities and suburbs
as the urban portions of the remaining Metropolitan Areas (MAs),
small towns as the urban portions of non-MAs, and rural areas as all
remaining portions of counties.
\20 Lief Jensen, Families in Poverty: Patterns, Contexts, and
Implications for Policy, Washington, D.C.: Family Impact Seminar,
1992.
CHILDREN OF SINGLE
PARENTS MORE LIKELY TO
PARTICIPATE
-------------------------------------------------------- Letter :3.2.3
Children in single-parent families are about 18 percent more likely
to attend preschool than those in married-couple families after
accounting for select demographic characteristics.\21 This percentage
difference is similar for poor children (21 percent). Research
suggests that single parents may use preschool as one form of child
care.
--------------------
\21 The characteristics we controlled for were race/ethnicity, family
income, immigrant status, linguistic-isolation status, work status of
parent, education status of the most educated parent, urbanicity of
residence, and state of residence.
SOME RISK FACTORS HAVE
LITTLE INFLUENCE
---------------------------------------------------------- Letter :3.3
After controlling for other characteristics, immigrant children, LI
children, and children with parents who do not work are nearly as
likely to attend preschool as their counterparts. While LI children
are about 33 percent less likely to attend preschool than their
non-LI counterparts, they are only 3 percent less likely after
controlling for other characteristics (see table 1.1).\22 Similarly,
the gap between immigrant and nonimmigrant closed from 22 percent to
10 percent\23 and the gap for children with working parents versus
nonworking parents closed from 29 percent to 10 percent after
controlling for other characteristics.\24
Table 1.1
Difference in Likelihood of
Participating in Preschool Relative to
Reference Group
--------------------
\22 The characteristics we controlled for were race/ethnicity, family
income and type, immigrant status, work status of parent, education
status of the most educated parent, urbanicity of residence, and
state of residence.
\23 The characteristics we controlled for were race/ethnicity, family
income and type, linguistic-isolation status, work status of parent,
education status of the most educated parent, urbanicity of
residence, and state of residence.
\24 The characteristics we controlled for were race/ethnicity, family
income and type, immigrant status, linguistic-isolation status,
education status of the most educated parent, urbanicity of
residence, and state of residence.