A Sixty Million Dollar Statistical Issue Arising in the Interpretation and Calculation of a Measure of Relative Disparity: Zuni Public School District 89 v. U.S. Department of Education

JOSEPH L. GASTWIRTH Department of Statistics, George Washington University, Washington DC 20052, USA (This paper has been accepted, subject to minor editing, by Law, Probability and Risk) In 1994, the U.S. Congress rewrote the statute providing educational funds to towns and cities of the nation where the federal government owns substantial amounts of land, which cannot be taxed locally. Usually, this impact aid goes to the local school districts; however, if a state has a program for subsidizing education in all areas so that the per-pupil expenditures are approximately equal throughout the state, then the state receives a large share of the federal funds. The previous statute authorized the Department of Education to establish a formula for determining whether a state’s educational funding was “equalized”. The new law specifies its own formula. Both formulas place an upper limit on a percentile-based measure of disparity calculated from data on average per-pupil expenditures in the local educational areas of the state. While the formulas are similar, they describe the percentiles to be used in the calculation differently. In 1996, when the Department of Education published its regulations for administering the new law it interpreted the new formula for assessing whether the educational funding in a state was “equal” to mean the identical calculation it had previously issued. In 1999 the state of New Mexico claimed that its school funding system met the criterion for “equalization” so it deserved 75% of the federal impact funds allocated to all impacted school districts in the state. Two local districts filed suit claiming that the state’s funding did not meet the new statute’s criterion for being “equalized” so the entire federal payment should go to the affected school districts. This article describes the calculation most statisticians would make after reading the new law. It will be seen that none of the calculations presented by the parties is technically correct. Furthermore, one of the assertions made by the Department of Education and accepted by the majority in a split appellate opinion is not mathematically correct. An en banc opinion of all twelve Circuit judges split on the interpretation of the statute and the U.S. Supreme Court has accepted the case for review. The last section suggests that alternative disparity measures based on data on expenditures in schools rather than districts would give a more accurate indication of the disparity in school funding in a state. Under either interpretation of the current law, a disparity in educational funding that courts previously have found to support a claim of racial segregation or discrimination, can satisfy the criteria for “equal” educational funding.