State-Funded Pre-Kindergarten: What the Evidence Shows. Endnotes


(1) Strengthening Head Start, What the Evidence Shows, U. S. Department of Health and Human Services, June 2003.

(2) The American Federation of Teachers, At the Starting Line, 2003. State pre-kindergarten was defined broadly as providing funds for preschool programs that have early childhood education or school readiness as a goal.

(3) Quality Counts 2002: Education Week, No. 17, January 10, 2002. Different surveys show slightly higher or lower numbers of states offering pre-kindergarten depending on the criteria that were used to define state pre-kindergarten. For example, to be included in some surveys the state must have designed, implemented and funded its own pre-kindergarten program while other surveys, such as the one conducted by the American Federation of Teachers, included states that fund pre-kindergarten through supplements to programs not administered by the state, such as Head Start.

(4) The number of states that give priority to low income children varies depending primarily on the definition of state-funded pre-kindergarten used in the survey. For example, The American Federation of Teachers publication, At the Starting Line, reports that 36 programs give priority to low income children, whereas Education Week, Quality Counts 2002 reports that 26 states do.

(5) Quality Counts 2002: Education Week, No. 17, January 10, 2002.

(6) Quality Counts 2002: Education Week, No. 17, January 10, 2002.

(7) Education Commission of the States, 2003 State of the State Addresses: Early Learning, February 2003 (updated March 17, 2003). Complied by Gloria Zradicka.

(8) Quality Counts 2002: Education Week, No. 17, January 10, 2002.

(9) Education Commission of the States. Pre-kindergarten Quick Facts, 2003.

(10) National Center for Education Statistics, U.S. Department of Education. Pre-kindergarten in U.S. Public Schools: 2000-2001. Statistical Analysis Report, March 2003.

(11) This figure includes teachers in the 15% of pre-kindergarten classes that were devoted exclusively to special education. The survey does not give teacher education levels for teachers in general education classrooms or combined general education and special education classrooms.

(12) Data for class size include general education classrooms or classrooms that combine general and special education. Data for classes that provide only special education are not included. The survey did not report teacher-child ratios.

(13) Research showing associations between children’s development and characteristics of program quality such as class size, teacher-child ratio, teacher education and qualifications, and parent involvement is reviewed in a recent report: National Research Council (2001). Eager to Learn: Educating Our Preschoolers (Chapter 4, pp. 128-181). B. T. Bowman, M. Donovan, & M.S. Burns (Eds.), Board on Children, Youth, and Families, Commission on Behavioral and Social Sciences Education Washington, DC: National Academy Press. Additional information can be found in National Research Council and Institute of Medicine (2000). From Neurons to Neighborhoods: The Science of Early Childhood Development (Chapter 11, pp. 314-317). J. P. Shonkoff and D. A Phillips (Eds.), Board on Children, Youth, and Families, Commission on Behavioral and Social Sciences and Education. Washington, D.C.: National Academy Press. The National Association for the Education of Young Children (NAEYC) and the Head Start Performance Standards require a class size with a maximum of 20 children and a teacher-child ratio of 1:10, which have become widely agreed upon standards. Research shows that smaller class sizes and teacher-child ratios are associated with better child outcomes, but research has not documented that a particular class size or teacher-child ratio is critical.

Though some studies show better children’s outcomes when parents are involved in their children’s education, there is little evidence to show what type of parent involvement is best and which type is best for particular types of children and families.

The National Academy of Sciences conducted a review of research on teacher quality and children’s outcomes for the 2000 report, Eager to Learn: Educating Our Preschoolers. The report recommends that early childhood teachers have bachelor’s degrees. Whether or not the available data adequately support the report’s conclusion and recommendations about the bachelor’s degree has been debated since the report was published. Most studies on teacher qualifications measure the association between the amount of teacher education and teacher quality or child outcomes. The results show that more education and training is associated with high quality teaching behaviors and positive child outcomes in a range of areas that include language, cognitive, social, and emotional development. These findings hold across Head Start, center-based child care, and child care homes (and for parent care). The few studies that have categorized teacher education into levels and types show that children whose teachers have at least an AA in early childhood education have better outcomes than children whose teachers have a Child Development Associate credential or high school degree. Children whose teachers have at least a bachelor’s degree attain the highest levels of competency. Some studies find that having a bachelor’s degree in a child-related field is associated with greater teacher competency and more positive child outcomes, whereas other studies find that having a bachelor’s degree regardless of the specialization is the better predictor of teachers’ performance and children’s verbal and cognitive skills.

Research on teacher qualifications and children’s outcomes has not addressed threshold levels for teacher education and training at which point further improvements do not yield additional benefits for children. The incremental change in child outcomes associated with each advance in education and training (i.e., moving from a CDA to an AA or from an AA to a BA) has not been studied. Randomized experiments have not been conducted in which children were randomly assigned to classrooms with teachers who differed in education levels and very few have controlled statistically for family or program characteristics that could affect child outcomes, such as maternal education, and other characteristics of the program or setting, such as level of pay and program administration support for advanced degrees. Very few studies have examined changes in children’s outcomes as their teachers acquire additional education and training across time.

The Child Development Associate (CDA) credential is awarded by the Council for Professional Recognition in Washington DC and funded by the Administration for Children, Youth, and Families in the U.S. Department of Health and Human Services to improve the quality of child care. This national credential program trains early care and education providers to promote children’s physical, social, emotional and intellectual growth in a center-based, home visitor or family child care program, and to work with parents and other adults to meet children’s needs in these areas.

(14) Gilliam W. & Ripple, C. (in press). What can be learned from state-funded prekindergarten initiatives? A data-based approach to the Head Start devolution debate. In E. Zigler & S.J. Styfco (Eds.), The Head Start debates (friendly and otherwise). New Haven, CT: Yale University Press.

(15) Bryant, D. M., Burchinal, M., Law, L. B., & Sparling, J. J. (1994). Family and classroom correlates of Head Start children’s developmental outcomes. Early Childhood Research Quarterly, 9, 289-309.

Burchinal, M. R., Cryer, D., Clifford, R. M., & Howes, C. (2002). Caregiver training and classroom quality. Applied Developmental Science, 6(1), 2-11.

Cassidy, D. J., Buell, M. J., Pugh-Hoese, S., & Russell, S. (1995). The effect of education on child care teachers’ beliefs and classroom quality: Year one evaluation of the teach early childhood associate degree scholarship program. Early Childhood Research Quarterly, 10, 171-183.

Dunn, L. (1993). Proximal and distal features of day care quality and children’s development. Early Childhood Research Quarterly, 8, 167-192.

Howes, C. (1997). Children’s experiences in center-based child care as a function of teacher background and adult:child ratio. Merrill-Palmer Quarterly, 43, 404-425.

Kontos, S., Hsu, H.C., & Dunn, L. (1994). Children’s cognitive and social competence in child care centers and family day care homes.Journal of Applied Developmental Psychology, 15, 87-111.

National Institute of Child Health and Human Development, Early Child Care Research Network (2000). Characteristics and quality of child care for toddlers and preschoolers. Applied Developmental Science, 4, 116-135.

Peisner-Feinberg, E. S., & Burchinal, M. R. (1997). Relations between preschool children’s child care experiences and concurrent development: The Cost, Quality, and Outcomes Study. Merrill-Palmer Quarterly, 43, 451-477.

Peisner-Feinberg, E. S., Burchinal, M. R., Clifford, R. M., Culkin, M. L., Howes, C., Kagan, S. L., & Hazejian, N. (2001). The relation of preschool quality to children’s cognitive and social developmental trajectories through second grade. Child Development, 72(5), 1534-1553.

Ruopp, R., Travers, J, Glantz, F. M, & Coelen, C. (1979). Children at the center: Summary findings and their implications. In Final report of the National Day Care Study: Children at the center, Vol. 1, Cambridge, MA: Abt Associates.

Whitebook, M., Howes, C., & Phillips, D. (1989). Who cares? Child care teachers and the quality of care in America: Final report of the National Child Care Staffing Study. Child Care Employee Project. Abt Associates.

(16) Ripple, C. H., Gilliam, W. S., Zigler, E., & Chanana, N. (1996). Will fifty cooks spoil the broth? The debate over entrusting Head Start to the states. American Psychologist, 54, 327-343.

(17) Gilliam & Ripple, in press.

(18) The District of Columbia is referred to as a state in the survey and in this report. This study identified fewer states as having pre-kindergarten programs, primarily because of the criteria that were used to define state pre-kindergarten. To be included in this study, the program: (a) must target or be accessible to children from low-income families, (b) provide at least some form of classroom-based, educational service directly to preschool-age children, (c) be implemented and administered at the state level (not state aid for low-income parents to purchase their own preschool services), (d) be primarily state-funded (not state supplementation to programs funded primarily at the federal or local level); and (e) not serve exclusively children with disability. Seven states (Alabama, Connecticut, Nevada, New Mexico, North Carolina, Pennsylvania, and Rhode Island) were not included in the survey but were included in the Education Week report, Quality Counts 2000, which used a broader definition of state-funded pre-kindergarten and documented 39 states and the District of Columbia as funding state pre-kindergarten.

(19) U.S. Department of Health and Human Services, Head Start Bureau. Head Start Program Information Report for the 2001-2002 Program Year, National Level Summary Report.

(20) Quality Counts 2002: Education Week, No. 17, January 10, 2002.

(21) American Federation of Teachers, At the Starting Line, 2003.

(22) The Education Commission of the States, State Funded Pre-k Profiles.

(23) The Abecedarian Project was a carefully controlled study of an early childhood intervention that began in the 1970’s. Fifty-seven infants from low-income families living in a small North Carolina town were randomly assigned to receive an intensive early intervention in a high quality child care setting and 54 were in a non-treated control group. Positive impacts were found on children’s cognitive development, school achievement, and long-term educational attainment. The findings have been published in peer review journals and vetted for technical soundness. The Perry Preschool Study, which began in the 1960s, was one of the first to identify lasting effects of high quality preschool programs on children’s outcomes. One hundred twenty-three poor African American 3- and 4-year-olds were randomly assigned either to attend a high quality preschool program or to no preschool. Both the Abecedarian Project and the Perry Preschool Study focused on specific approaches to offering early childhood education services: neither focused on state-funded pre-kindergarten.

For more information about the Abecedarian Project study methods and results, see:

Campbell, F. A., Ramey, C.T., Pungello, E. P., Sparling, J., & Miller-Johnson, S. (in press). Early childhood education: Young adult outcomes from the Abecedarian Project. Applied Developmental Science.

Campbell, F. A., Punbello, E. P., Miller-Johnson, S., Burchinal, M., & Ramey, C.T. (in press). The development of cognitive and academic abilities: Growth curves from an early childhood educational experiment. Developmental Psychology.

Ramey, C.T., Campbell, F. A., Burchinal, M., Skinner, M. L., Gardner, D.M., & Ramey, S. L. (2002). Persistent effects of early intervention on high-risk children and their mothers. Applied Developmental Science, 4, 2-14.

For more information about the Perry Preschool Study methods and results, see:

Schweinhart, L.J., Barnes, H. V., & Weikart, D. P. (1993). Significant benefits: The High/Scope Perry Preschool study through age 27. Monographs of the High/Scope Educational Research Foundation, 10. Ypsilanti, MI: High/Scope Press.

(24) Studies of the Chicago Child-Parent Centers do not include randomized designs.

(25) New Jersey Department of Education, Office of Early Childhood Education. Abbott Preschool Program Implementation Guidelines, February 2003.

(26) New York State Education Department. Universal PreKindergarten Takes Off in New York State, February 2000.

(27) Illinois Office of the Governor. Ready, Set, Grow: A Framework for Universal Access to Quality Preschool in Illinois Created by the Governor’s Task Force on Universal Access to Preschool

(28) Gilliam, W. & Zigler, E. (2001). A critical meta-analysis of all evaluations of state-funded preschool from 1977 to 1998: Implications for policy, service delivery and program evaluation. Early Childhood Research Quarterly, 15; Gilliam & Ripple, in press.

(29) Since 1998, state-funded pre-kindergarten programs have evolved and may not reflect the current-day program. For example, the New York program included in the 10-state review, the New York State Experimental Prekindergarten Program (NYSEPP), is an older program implemented in 1966 and not New York’s current universal pre-kindergarten program which was implemented in 1998 and which has not been formally evaluated.

(30) Gilliam & Zigler, 2001; Gilliam & Ripple, in press.

(31) The term cohort refers to a group of children who entered and exited the program the same year.

(32) Of the 10 studies, Gilliam and Zigler note that only two have been published in professional journals, which require a peer review process to evaluate technical soundness.

(33) Gilliam & Zigler, 2001.

(34) Gilliam & Zigler, 2001.

(35) Gilliam & Zigler, 2001.

(36) Gilliam & Zigler, 2001.

(37) Gilliam & Zigler, 2001.

(38) Gilliam & Zigler, 2001.

(39) Gilliam & Ripple, in press.

(40) Gilliam & Ripple, in press.

(41) Private school is defined as schools or child care centers that offered educational and developmental programs for four-year-olds in exchange for tuition or fees.

(42) Standardized assessments included measures of receptive vocabulary, expressive language skills, cognition and pre-mathematics, and letter and word recognition. Direct assessments used in other large-scale studies measured “basic skills” including children’s ability to recognize colors, recognize written numerals, and understand story and print concepts. (These tests do not have national norms.) Teachers rated children’s conversation, communication and expressive language skills as well as the appropriateness of children’s classroom behaviors, which were used as the measure of social and emotional development. Children’s health and physical well-being were assessed through teacher and parent surveys. (Teacher ratings and teacher and parent surveys also do not have national norms.)

(43) This method is superior to computing the amount of change in children’s scores between a pre-test prior to intervention and a single post-test following intervention for several technical reasons. For example, the technical requirements for this method of analysis, which includes that children’s abilities be measured at three or more time points, significantly reduce measurement error allowing stronger inferences to be made about the amount and direction of children’s progress. Process quality refers to interactions with teachers and peers, classroom activities, resources and materials, and instructional methods that children experience in the program.

(44) The Early Childhood Environment Rating Scale-Revised (ECERS-R) was used to measure classroom quality. Areas evaluated include: Space and Furnishings (e.g., room arrangement for play), Personal Care Routines (e.g., health and safety practices), Language-Reasoning (e.g., presence and use of books and pictures, encouraging children to use language), Activities (e.g., art, music, blocks, dramatic play, numbers), Interaction (discipline, peer interactions, sensitivity and responsiveness of staff-child interactions), Program Structure (e.g., free play, group time), and Parents and Staff (e.g., provisions for personal needs of staff, opportunities for professional growth). The measure was not designed to evaluate the range of specific activities and practices that have been shown in research to promote knowledge and skills in the areas of language, early literacy, early mathematics, and social and emotional competency.

(45) The failure to find an association between letter-word recognition and program quality is not surprising because although the program quality measure used in this study is technically sound and widely used, it does not have a focus on letter-word recognition or other pre-reading knowledge and skills. Program quality measures that include observations of teacher-child interactions and activities that promote pre-reading skills have not been sufficiently developed.

(46) Georgia’s pre-kindergarten program had significantly fewer black children and significantly more white children than Head Start. Compared to Head Start, Georgia’s pre-kindergarten program had fewer children whose mothers or fathers had less than a high school degree, fewer children receiving Medicaid or SCHIP as their primary insurance, fewer children whose families received welfare during the past five years, fewer children who were born to teenage mothers, fewer children who lived in a household with less than two adults and fewer children who had not lived continuously with both parents from birth, and fewer children whose families had received food stamps during the past five years. Average household income was significantly higher for Georgia pre-kindergarten than for Head Start (40,000-50,000 versus 20,000-30,000). Georgia pre-kindergarten had more children whose mothers had a bachelor’s degree, more children who were read to every day at home, and more parents participating in their children’s education. In addition, more children in Head Start were referred for language services. Children in Georgia’s pre-kindergarten program also began the pre-kindergarten program with scores above the national norm for letter-word recognition, whereas Head Start scored below national norms. Though children entering both programs scored below national norms on all vocabulary, expressive language, story comprehension and print familiarity, and problem-solving/pre-mathematics, children in Head Start had entering scores that were substantially lower than children in Georgia pre-kindergarten, as well as lower scores on a non-standardized assessment of basic skills mastery (color recognition, number recognition, and counting).

Though Georgia pre-kindergarten differed from private preschools on some characteristics, they were similar in most areas. Georgia’s pre-kindergarten program had significantly more black children and significantly fewer white children than private preschools. Compared to private preschools, Georgia’s pre-kindergarten program had the same percentages of children whose mothers or fathers had less than a high school degree, received Medicaid or SCHIP as their primary insurance, received welfare during the past five years, children born to teenage mothers, children who have lived with both parents continuously since birth, and children whose families had received food stamps during the past five years. Compared to private preschool, Georgia pre-kindergarten had the same percentage of children whose mothers had a bachelor’s degree, the same percentage of children who were read to every day at home, and the same percentage of children whose parents participated in their children’s education. In addition, the same percentages of children were referred for language services in Georgia pre-kindergarten and private preschool.

Children in Georgia’s pre-kindergarten program began the pre-kindergarten program with letter-word recognition scores that were lower than those for children in private preschool, though they were above the national norm. Children entering Georgia pre-kindergarten had vocabulary and expressive language scores that were significantly lower than children entering private preschools, though scores for both groups were below national norms. Children entering Georgia pre-kindergarten had problem-solving/pre-mathematics scores that were significantly lower than children entering private preschool, and below national norms. Scores for story comprehension and print familiarity were similar for children entering Georgia pre-kindergarten and public preschools. On non-standardized assessments of basic skills mastery (color recognition, number recognition, and counting) children entering Georgia pre-kindergarten scored significantly lower on number recognition and had similar scores for color recognition and counting, compared to children entering private preschools.

(47) Self-selection bias occurs because study participants with particular characteristics (such as extreme poverty) choose to belong to a specific group (such as Head Start) and those same characteristics also have a known or suspected effect on the outcome of interest (such as the effect of extreme poverty on children’s development and school readiness).

(48) In the sub-study, a portion of the larger sample of children who attended Georgia pre-kindergarten was selected to obtain a group of children with characteristics similar to those who attended Georgia Head Start. The groups were matched to ensure that children in each group were similar on geographical location (rural, urban, other), sex and race. The technique also tries to account for other family risk factors using a “propensity score,” which represents the probability that all of the children included in the sub-study have characteristics comparable to the children who enrolled in Georgia Head Start. More specifically, the propensity scores estimated the probability that all children in the sub-study belonged to one of four characteristic groups that tended to enroll in Head Start: 1) TANF eligible children whose primary caregiver generally had not completed high school, 2) low-income children eligible for food stamps and Medicaid whose primary caregiver generally had a high school education, 3) lower to middle-income children, and 4) children who had lived with both parents since birth and had parents whom teachers identified as participating in the classroom.

(49) The evaluation includes many comparisons between Georgia pre-kindergarten and Head Start. Most did not control for child and family differences that could account for program group differences. The sub-study used statistical controls for the influence of child and family characteristics on children’s outcomes and focused on disadvantaged groups of children who attended Georgia pre-kindergarten and Head Start. Therefore, when comparing Georgia pre-kindergarten and Head Start, this report summarizes only those analyses that attempted to control for child and family differences between the two groups.

(50) See Gilliam & Zigler for a more detailed discussion of the technical limitations of state-funded pre-kindergarten program evaluations.

(51) Bond, J. T. Interim Report: The Evaluation of Connecticut’s School-Readiness Program. Cohorts 1 & 2 through Spring 2002. Prepared for the Connecticut Department of Social Services and Connecticut Department of Education.

(52) The School Readiness Initiative in South Central Connecticut: Classroom Quality, Teacher Training, and Service Provision, FY 2000 Update. Report prepared by W. S. Gilliam, Yale University Child Study Center, January 2001.

(53) Bond, 2002

(54) The amount of change from fall to spring rose slightly, but did not reach the conventional level of statistical significance. This finding indicates that children were close to national norms when they entered pre-kindergarten, and that the pre-kindergarten program did not accelerate children’s advancement beyond the typical rate of development. Effect sizes for the pre-test/post-test difference are not reported and cannot be computed from the data provided.

(55) The report does not show children’s performance at the beginning and end of pre-kindergarten, as was done for school readiness. As a result, it is not known how close children were to national norms before entering the program and the amount of progress children made during the pre-kindergarten year.

(56) Investing in Better Outcomes: The Delaware Early Childhood Longitudinal Study. University of Delaware Center for Disability Studies, April 2002.

Henry, G. T. Gordon, G. S., Mashburn, A., & Ponder, B. D. (2001). Pre-K Longitudinal Study: Findings from the 1999-2000 School Year. Report prepared by the Applied Research Center, Andrew Young School of Policy Studies, Georgia State University for the Georgia Office of School Readiness.

Opfer, V. D., Brackett, M. H. H., & Henry, G. T. (1999). Impact of Georgia’s Pre-K Program on Kindergarten through Third Grade. Prepared for the Council for School Performance.

Maryland State Department of Education (March 2003). Children Entering School Ready to Learn: School Readiness Information. School Year 2002-2003 by State and County.

Illinois State Board of Education Research Division (June 2001). Illinois Prekindergarten program for children at risk of academic failure. FY 2000 Evaluation Report.

(57) Center for Law and Social Policy (CLASP). State Initiatives to Promote Early Learning: Next Steps in Coordinating Subsidized Child Care, Head Start, and State Pre-kindergarten. Policy brief, April 2001.

(58) National Association of State Boards of Education, Early Childhood Education Network. Creating a Unified, Comprehensive System of Early Childhood Education, 2003.

(59) IDEA is a federal formula grant program that assists states in providing a free and appropriate public education for children with disabilities in the least restrictive environment possible.

(60) Part B of IDEA assists states in providing a free and appropriate public education for individuals with disabilities from age 3 through 21. Part C supports early intervention services for infants and toddlers with disabilities from birth through age two and their families.

(61) Trust for Early Education website,

(62) Pew Charitable Trusts. Starting Early Starting Strong (2001).

(63) National Governors Association Website,
National Conference of State Legislatures,

(64) National Center for Children in Poverty. Map and Track 2000: State Initiatives for Young Children and Families. Mailman School of Public Health, Columbia University.

(65) California Children & Families Commission. First 5 California. Annual Report, Fiscal Year 2001/02.

(66) Kentucky Governor’s Office of Early Child Development, KIDS (Kentucky Invests in Developing Success) NOW

(67) Smart Start

(68) Smart Start and Preschool Child Care Quality in NC: Change Over Time and Relation to Children’s Readiness, March 2003. A report by the Frank Porter Graham Child Development Institute Smart Start Evaluation Team: University of North Carolina-Chapel Hill.

(69) The study design was quasi-experimental and involved pre-post assessments with control groups. Because the studies were not randomized experiments, changes in child outcomes cannot be attributed unequivocally to Smart Start. Researchers have had difficulty conducting randomized trials because a randomized design is viewed as antithetical to the state’s mission of providing essential services to all children in the state who need them. There is uncertainty about whether or not control group children had attended child care centers that offered some Smart Start services. In addition, length of participation varied greatly in the Smart Start group, ranging from 8 to 60 months, limiting conclusions about how much participation is needed to realize benefits. Data are not available on the duration and intensity of services needed to achieve benefits or which combination of services may contribute to positive effects and for which children.

(70) These findings are sometimes misinterpreted as showing that improvements in child outcomes are the direct result of quality improvement activities. However, the results cannot address this issue because Smart Start programs that focused on quality improvement were not directly compared to those that did not.

(71) North Carolina’s Early Care and Education System: Report to the North Carolina General Assembly, January 2003.

(72) South Carolina Office of Early Childhood Education,

(73) First Steps and Further Steps: Early Outcomes and Lessons Learned from South Carolina’s School Readiness Initiative, 1999-2002 Program Evaluation Report. Prepared by Child Trends, Inc.

(74) National Center for Children in Poverty. Map and Track 2000: State Initiatives for Young Children and Families. Mailman School of Public Health, Columbia University.

View full report


"paper.pdf" (pdf, 782.83Kb)

Note: Documents in PDF format require the Adobe Acrobat Reader®. If you experience problems with PDF documents, please download the latest version of the Reader®