Ordering birth certificate

ABBREVIATIONS. DOE, Department of Education; IDEA, Individuals With Disabilities Education Act.

States and public school districts are required by federal law to provide educational services for children with disabilities. This mandate can have significant financial implications, as nearly 6 million students--1 of 8--require special education classes. (1) For the 1999-2000 school year, states spent ~$50 billion on special education services. (2) The total excess cost associated with assigning students to special education is estimated to be 2.3 times higher on average than the cost of students with no special needs. (3) These additional expenditures entail difficult financial decisions. The federal government covers ~7.5% of the total additional expenditure for special education students. States must either raise taxes or reduce other expenditures to fund the mandated special education costs. Increases in special education placement often lead to reductions in spending on regular-education students. (4)

Although sociodemographic factors at birth have been found to be the most important predictor of placement in special education, adverse perinatal conditions also play a significant role. (5-8) This study builds on earlier analyses of special education placement of children in Florida elementary schools (5,9,10) and quantifies the fiscal implications of maternal and infant conditions at birth on state education costs.

In addition to special education, repeating a grade is a contributor to excess kindergarten costs. Although grade retention decisions are more discretionary than are special education placements, these local decisions have fiscal bearing on the state as well as on the school district that makes the retention decision. (11) Given the recent push at both the state and national levels to reduce the prevalence of social promotion, (12,13) grade retention may be a major determinant of excess costs and needs to be considered alongside special education placement costs. (14)

To date, little research has been conducted on the relationship between perinatal conditions and educational costs. Previous studies have estimated a total cost associated with a particular birth condition, usually low birth weight or preterm birth; these studies combine expenditures on health care, child care, and special education (15-17) Reviews of these economic impact studies have not distinguished proximate costs (extended neonatal intensive care unit stays) from distal costs (reduced earning capacity). (18,19) Little is known about the earliest determinants of excess educational costs. One US study analyzed a single determinant of special educational costs, low birth weight. (20) The authors used the 1988 National Health Interview Survey data set, which, although representative of the then-current US low birth weight rate of 7.5%, was limited to <8000 students aged 6 to 15. The present study differs from this earlier study of special education costs in several respects: 1) the present study considers the effects of a wider array of biomedical and sociodemographic conditions on excess educational costs; 2) cost estimates include the additional cost associated with grade retention, which tends to parallel average special education costs; 3) it examines an entire state's population so that comparisons may be made to other state education systems; 4) a larger sample size permits attention to be focused on rare but extremely costly special education conditions and more adverse birth conditions as well as more commonplace ones; and 5) it uses more current data and therefore is more reflective of current finance allocation decisions faced by states.

The purpose of the present study was to estimate state expenditures at kindergarten from infant and maternal medical and sociodemographic factors that are known at birth. This analysis is limited to kindergarten costs because of the desire to use the most current data possible, and kindergarten special education outcomes are highly correlated with later special education outcomes. In so doing, it seeks to enlarge understanding of the fiscal implications of adverse maternal and infant birth conditions at entry to public school.

METHODS

Children who were born in Florida between September 1, 1990, and August 31, 1991, would have been expected to enter kindergarten in the 1996-1997 academic year. However, to capture children who were enrolled electively in kindergarten 1 year later than expected as well as children who were required to repeat kindergarten, we examined all kindergarten records between 1996 and 1999.

The birth cohort was generated from records in the Florida Department of Health Vital Statistics database. Birth records were linked to school records obtained from the Florida Department of Education (DOE) databases for the relevant academic years. Of the 197 659 infants who were born in Florida during 1990-1991, 125 430 (63.5%) were matched successfully with DOE records.

All records that had any missing values on the variables of interest were eliminated from the matched sample (N = 4897). To ascertain the representativeness of the matched birth-to-school cohort, we compared the distribution of 12 predictor variables in the 1996-1999 kindergarten study sample with the total vital statistics 1990-1991 birth cohort (Table 1). The kindergarten sample was not substantively different from the total birth population, with 1 exception: the school-age sample had a higher proportion of children who were poor at birth compared with the original birth cohort (37.1% vs 30.6%). This discrepancy arises from the fact that low-income families are less likely to send their children to private schools or to leave the state. (21) Nevertheless, because of the potential for selection bias, all estimations for poverty and non-poverty subpopulations were repeated to ensure that the central findings were not sensitive to the overrepresentation of poverty families in the analysis data set. In addition, a set of cost-estimate simulations was conducted using study and population means for each predictor variable.

Variables

Outcome Variable

The dependent variable was set to the state expenditure on the student through his or her completion of kindergarten, expressed in 2001-2002 dollars. Costs were measured using state school finance system formulas. Costs were derived from the student's primary exceptionality code in the student's Federal/State Indicator record in the DOE database. A child's primary exceptionality code identifies the disability requiring the greatest allocation of personnel resources in cases in which >1 disability is diagnosed. Each exceptionality generates a different amount of state funding to the school district. Students who were retained in grade and who also had an exceptionality were counted twice, once in each category. Retained students who were classified with 1 exceptionality in their first year of kindergarten and who were classified with a different exceptionality in their second year of kindergarten were coded with the weights (cost multipliers) associated with each exceptionality.

Predictor Variables

Infants were classified into 4 mutually exclusive birth weight groups: <1000 g, 1000 to 1499 g, 1500 to 2499 g, and [greater than or equal to] 2500 g). Congenital anomaly and gender were dichotomous dummy variables. Children were divided into the racial/ethnic categories of white, black, Hispanic, and other (predominately Asian but including those identified as mixed race or Native American). Two time points for poverty were used. Poverty at birth was determined by verifying mother's eligibility for Medicaid during pregnancy. Poverty at kindergarten was determined by verifying child's participation in the school's free or reduced price lunch program. Eligibility for free or reduced price lunch is based on both family income and family size and like Medicaid-funded pregnancy services includes families whose income is <185% of the federal poverty level.