The Determinants of Per-Pupil Expenditures in Private Elementary and Secondary Schools: An Exploratory Analysis.

Although policy makers have expressed interest in expenditures by private elementary and secondary schools, nationally representative data in this area are not available. In the absence of expenditure data it may nevertheless be possible to derive national estimates of total expenditures by private schools indirectly, by using the information that some associations of private schools collect from their members. Data on expenditures are routinely collected by three major associations of private schools: the National Catholic Education Association, the National Association of Independent Schools, and the Lutheran Church--Missouri Synod. This paper extends an earlier study of their data by examining whether more accurate national expenditure estimates can be derived by linking expenditure data with information on school characteristics collected as part of the National Center for Education Statistics Schools and Staffing Survey (SASS). If an appropriate model relating school characteristics and per-pupil expenditures could be estimated, it might be used to impute per-pupil expenditures for the full sample of private schools. Models incorporating a number of variables were explored, but in spite of the researchers' success in identifying school characteristics related to per-pupil expenditures, these characteristics do not fully explain the large differences in per-pupil expenditures among three sectors of private schools in this study: Lutheran day schools, independent day schools, and boarding schools. It is suggested that the model for Lutheran day schools is likely to provide reasonable estimates of per-pupil expenditures for many religious and nonsectarian private schools, but this assumption cannot be checked with the data available. To obtain improved estimates of the total amount spent by private schools in the United States, it will be necessary to collect data from a national sample of private schools, either through the SASS or a new special-purpose survey. An appendix contains 4 tables reporting results from the models predicting various school characteristics. (Contains 12 tables and 4 references.) (SLD)