P, school districts in New Mexico, challenged a determination by D that the state had equalized spending in its school districts. The Federal Impact Aid Program provides financial assistance to local school districts whose ability to finance public school education is adversely affected by a federal presence. The statute prohibits a State from offsetting this federal aid by reducing state aid to a local district. The statute contains an exception permitting a State to reduce its own local funding on account of the federal aid where D finds that the state program 'equalizes expenditures' among local school districts. 20 U.S.C. § 7709 (b)(1). D is required to use a formula that compares the local school district with the greatest per-pupil expenditures in a State to the school district with the smallest per-pupil expenditures. If the former does not exceed the latter by more than 25 percent, the state program qualifies as one that 'equalizes expenditures.' To do this D must 'disregard districts with per-pupil expenditures above the 95th percentile or below the 5th percentile. Regulations require that D create a list of school districts ranked in order of per-pupil expenditure; then identify the percentile cutoff point and compare the highest spending and lowest spending of the remaining school districts to see whether they satisfy the statute's requirement that the disparity between them not exceed 25 percent. D ranked all of New Mexico’s 89 school districts. D listed each of New Mexico's 89 local school districts in order of per-pupil spending for fiscal year 1998. After ranking the districts, d excluded 17 school districts at the top of the list because those districts contained less than 5 percent of the student population; for the same reason, they excluded an additional 6 school districts at the bottom of the list. The 66 districts left accounted for 90 percent of the State's student population. The 66 districts left accounted for 90 percent of the State's student population. Of those, the highest ranked district spent $3,259 per student; the lowest ranked district spent $2,848 per student. The difference of spending per pupil was $411. This was less than 25 percent of the lowest per-pupil figure. D then qualified the state program as a program that 'equalizes expenditures.' New Mexico was, therefore, free to offset federal impact aid to individual districts by reducing state aid to those districts. Ps argued that the regulations are inconsistent with the authorizing statute. Ps claim the statute requires D to calculate the 95th and 5th percentile cutoffs solely on the basis of the number of school districts (ranked by their per-pupil expenditures), without any consideration of the number of pupils in those districts. Only 10 districts (accounting for less than 2 percent of all students) would have been identified as the outliers). The highest and lowest per-pupil expenditures of the remaining districts (26.9 percent) would exceed 25 percent. An administrative law judge rejected Ps’ challenge. Ps appealed to the Court of Appeals, which affirmed. En banc, the court affirmed again. The Supreme Court granted certiorari.