View PDF Version

No. 05-1508

In the Supreme Court of the United States

ZUNI PUBLIC SCHOOL DISTRICT NO. 89, ET AL., PETITIONERS

v.

DEPARTMENT OF EDUCATION, ET AL.

ON WRIT OF CERTIORARI
TO THE UNITED STATES COURT OF APPEALS
FOR THE TENTH CIRCUIT

BRIEF FOR THE FEDERAL RESPONDENT

PAUL D. CLEMENT
Solicitor General
Counsel of Record

PETER D. KEISLER
Assistant Attorney General

EDWIN S. KNEEDLER
Deputy Solicitor General

SRI SRINIVASAN
Assistant to the Solicitor
General

ROBERT S. GREENSPAN
PETER R. MAIER
Attorneys
Department of Justice
Washington, D.C. 20530-0001
(202) 514-2217

 

KENT D. TALBERT
General Counsel

STEPHEN H. FREID
MARK SMITH
Attorneys
Department of Education
Washington, D.C. 20202

QUESTION PRESENTED

The federal Impact Aid statute, 20 U.S.C. 7701 et seq., provides funds to local school districts that have a substantial federal presence within the district. The Impact Aid program generally prohibits a State from considering federal impact aid funds received by local school districts when allocating state funds among school districts in the State. 20 U.S.C. 7709(a). If, how ever, the Secretary of Education certifies that the State's funding system is "equalize[d]" within the meaning of the statute, 20 U.S.C. 7709(b) (2000 & Supp. III 2003), the State may consider federal impact aid funds received by a school district when allocating state funds among school districts. The question presented is as follows:

Whether the methodology used by the Secretary in determining whether a State's funding system is equalized is based on a permissible interpretation of the Impact Aid statute, 20 U.S.C. 7709(b) (2000 & Supp. III 2003).

In the Supreme Court of the United States

No. 05-1508

ZUNI PUBLIC SCHOOL DISTRICT NO. 89, ET AL., PETITIONERS

v.

DEPARTMENT OF EDUCATION, ET AL.

ON WRIT OF CERTIORARI
TO THE UNITED STATES COURT OF APPEALS
FOR THE TENTH CIRCUIT

BRIEF FOR THE FEDERAL RESPONDENT

OPINIONS BELOW

The per curiam opinion of the equally divided en banc court of appeals (Pet. App. 1a-2a) is reported at 437 F.3d 1289. The panel opinion of the court of appeals (Pet. App. 3a-33a) is reported at 393 F.3d 1158. The decision of the Secretary of Education (Pet. App. 34a- 40a) is unreported.

JURISDICTION

The judgment of the court of appeals was entered on February 23, 2006. The petition for a writ of certiorari was filed on May 24, 2006, and was granted on Septem ber 26, 2006. The jurisdiction of this Court rests on 28 U.S.C. 1254(1).

STATUTORY AND REGULATORY
PROVISIONS INVOLVED

The pertinent statutory and regulatory provisions are set forth in an appendix to this brief. App., infra, 1a-8a.

STATEMENT

1. a. Congress established the Impact Aid program in 1950. 20 U.S.C. 7701 et seq.; Act of Sept. 30, 1950, ch. 1124, § 4(c)-(e), 64 Stat. 1106. The program provides federal funds to local school districts-or "local educa tional agencies" (LEAs), see 20 U.S.C. 7713(9)-to assist with the financial burdens of providing "educational ser vices to federally connected children." 20 U.S.C. 7701. "Federally connected children" include children whose parents are in the military services or are civilian fed eral employees, children who reside on Indian lands or federal property, and children whose parents are em ployed on federal property. See 20 U.S.C. 7701(2)-(5). The Impact Aid program also provides funds to school districts whose property tax base is adversely affected by the federal government's tax-exempt ownership of real property in the districts. See 20 U.S.C. 7701(1).

As originally enacted, the Impact Aid law did not address whether a State, when allocating state educa tional funding among LEAs, could take into account the extent to which a particular LEA was the recipient of federal Impact Aid funds. In 1968, Congress specifically addressed that issue "in response to court decisions which held that some States, in considering the amount of impact aid payments received by a school district, and in deducting that amount from the district's State aid payment, were unfairly penalizing districts because of the existence of the Federal payments." S. Rep. No. 763, 93d Cong., 2d Sess. 55 (1974). Congress accordingly amended the Impact Aid statute to prohibit States from taking into account an LEA's receipt of Impact Aid funds when determining the allocation of state education funding among LEAs. See 20 U.S.C. 7709(a).

b. In the 1970s, an increasing number of States- often in response to legal challenges to their education funding systems-began revising their funding pro grams in an effort to reduce disparities and promote equalization in funding among school districts. See S. Rep. No. 763, supra, at 55; Staff of the House Comm. on Educ. and Labor, 93d Cong., 2d Sess., Public Law 874 and State Equalization Plans: The Problems of the Leg islative Prohibition of Section 5(d)(2), at 1, 12, 14 (Comm. Print 1974); GAO, School Finance: State Efforts to Equalize Funding Gaps Between Wealthy and Poor Districts 15, 21 (1998) (1998 GAO Report). In light of the "move by a number of jurisdictions to reform their school finance programs to assure equalization of educa tional expenditures," Congress became concerned that the "[i]nability to consider impact aid payments for the purposes of establishing an equalized level of expendi ture seriously interfered with State plans for school fi nance reform." S. Rep. No. 763, supra, at 55; see H.R. Rep. No. 805, 93d Cong., 2d Sess. 42 (1974). Congress recognized that, when a State attempts to promote equalization of funding among LEAs, barring the State from considering impact aid to individual LEAs would impair the State's ability to advance its equalization ob jective. See ibid.

Congress therefore amended the Impact Aid statute to permit a State to take into account an LEA's receipt of funds under the program when allocating state fund ing if the "State has in effect a program of State aid that equalizes expenditures for free public education among [LEAs] in the State." 20 U.S.C. 7709(b)(1). Congress left the task of formulating criteria for determining whether a State operates a qualifying equalization pro gram to the discretion of the Secretary of Education. See 20 U.S.C. 240(d)(2)(B) (Supp. IV 1974) (providing that the term "equalize expenditures" would "be defined by the [Secretary] by regulation"). In 1976, after notice and comment, the Secretary promulgated regulations establishing the rules for determining whether a State has an effective equalization program such that it may consider Impact Aid funding when distributing state funds. See 41 Fed. Reg. 26,320, 26,327 (1976). Under those rules, a State would be considered "equalized" if the disparity in per-pupil expenditures or revenues among the State's LEAs was no more than 25%. Ibid.

In an appendix to the regulations, the Secretary set forth the methodology for applying the 25% disparity criterion. 41 Fed. Reg. at 26,329. First, the Secretary ranked the State's LEAs according to each LEA's per- pupil expenditures or revenues. Ibid. Next, the Secre tary incorporated the number of pupils in each school district in order to identify the per-pupil expenditure or revenue for the LEA that serves the pupils at the 95th and 5th percentiles of all pupils in the State, as arrayed on a per-pupil revenue or expenditure basis. Ibid. Then, the Secretary compared the per-pupil expendi tures or revenues of those two LEAs to determine whether the disparity exceeded 25%. Ibid. In short, as the Secretary explained, "[i]f there is a disparity of no more than 25 per cent in revenues per pupil * * * available to the 95th and 5th percentile school districts (those with the 95th and 5th percentiles of the total number of pupils after being ranked in order of revenue per pupil), the program would be deemed to qualify" as equalized. Id. at 26,320.

The effect of comparing "the 95th and 5th percentile school districts" and incorporating the overall pupil pop ulations into the calculus was essentially to exclude those LEAs whose pupil populations accounted for roughly that 5% of the overall pupil population in the State that lay at the far ends of the spectrum of per-pu pil expenditures or revenues, and correspondingly to base the disparity calculation on the per-pupil expendi tures or revenues associated with the central 90% of the overall pupil population along the spectrum. The Secre tary explained that the "exclusion of the upper and bot tom 5 percentile school districts is based upon the ac cepted principle of statistical evaluation that such per centiles usually represent unique or noncharacteristic situations." 41 Fed. Reg. at 26,320.

In the course of the notice and comment rulemaking process, the Secretary specifically rejected the sugges tion (embraced by petitioners here) that the Secretary should not account for the number of pupils in each LEA, but should just compare the 95th and 5th percen tiles of LEAs instead of the LEAs associated with the 95th and 5th percentiles of pupils. The Secretary ac cordingly prescribed that the "percentiles will be deter mined on the basis of numbers of pupils and not on the basis of numbers of districts." 41 Fed. Reg. at 26,324. The Secretary was concerned that, failing to account for the number of pupils in each school district could skew the calculation if, for example, very large LEAs-accounting for a significant percentage of the State's pupils-lay at each end of the spectrum. The Secretary explained:

[B]asing an exclusion on numbers of districts would act to apply the disparity standard in an unfair and inconsistent manner among States. * * * In States with a small number of large districts, an exclusion based on percentage of school districts might exclude from the measure of disparity a substantial percent age of the pupil population in those States. Con versely, in States with large numbers of small dis tricts, such an approach might exclude only an insig nificant fraction of the pupil population and would not exclude anomalous characteristics.

Ibid. In 1993, the 1976 regulations and appendix were codi fied at 34 C.F.R. 222.63 (1993) and 34 C.F.R. Pt. 222, Subpt. K, App. (1993).

c. Until 1994, the methodology for determining whether a State operates an effective equalization stan dard was set forth exclusively in the regulations, with no statutory provisions addressing the matter. In connec tion with the periodic reauthorization of the Impact Aid program to take place that year, the Secretary submit ted to Congress comprehensive proposed legislation to reauthorize the program. See S. 1513, 103d Cong., 1st Sess. (1993); H.R. 3130, 103d Cong., 1st Sess. (1993); see also S. Rep. No. 292, 103d Cong., 2d Sess. 56 (1994) (observing that "S. 1513 as introduced was a direct transmission from the Administration"). As one aspect of the Secretary's legislative proposal, the Secretary proposed language that would for the first time codify in the Impact Aid statute the standards for determining whether a State operates a qualifying equalization pro gram.

The Secretary's proposal retained the 25% disparity standard set forth in the regulations, subject to the 95th percentile exclusion. See S. 1513, supra, at 352 (pro posed §8009(b)(2)(B)(i)); H.R. 3130, supra, at 352 (same). With particular respect to the 95th percentile exclusion, the Secretary's proposed language provided for "disregard[ing] [LEAs] with per-pupil expenditures or revenues above the 95th percentile of such expendi tures or revenues in the State." Ibid. Congress en acted, without change, the language proposed by the Secretary. The sole change from the Secretary's pro posal was to retain an exclusion at the bottom of the range for the 5th percentile. Congress, however, made no change with respect to the Secretary's language de scribing the method for determining which LEA falls at the 95th percentile (or the 5th percentile).1

As enacted, the 1994 amendment specifies that a State will be considered to operate a program that "equalizes expenditures among local educational agen cies"-and thus will be permitted to consider federal Impact Aid funding when distributing State aid-if the "Secretary determines" and "certifies" that:

the amount of per-pupil expenditures made by, or per-pupil revenues available to, the [LEA] in the State with the highest per-pupil expenditures or rev enues did not exceed the amount of such per-pupil expenditures made by, or per-pupil revenues avail able to, the [LEA] in the State with the lowest such expenditures or revenues by more than 25 percent.

20 U.S.C. 7709(b)(1) and (2)(A). The statute then states, con sistent with the language proposed by the Secretary, that "[i]n making [that] determination, * * * the Secretary shall-"

disregard [LEAs] with per-pupil expenditures or revenues above the 95th percentile or below the 5th percentile of such expenditures or revenues in the State.

20 U.S.C. 7709(b)(2)(B)(i).

In the wake of the 1994 amendments to the Impact Aid law, the Secretary issued new regulations. With respect to the 25% disparity standard and the 95th and 5th percentile exclusions, the regulations essential- ly mirrored the statutory language. See 34 C.F.R. 222.162(a). In particular, the regulations stated that a State will be considered to operate a program that "equalizes expenditures if the disparity in the amount of current expenditures or revenues per pupil for free pub lic education among LEAs in the State is no more than 25 percent." 34 C.F.R. 222.162. The regulations further provided that, "[i]n determining the disparity percent age, the Secretary disregards LEAs with per pupil ex penditures or revenues above the 95th or below the 5th percentile of those expenditures or revenues in the State." Ibid. In an appendix to the regulations, the Sec retary set forth the precise methodology for applying the disparity test, which retained the same methodology for implementing the 95th and 5th percentile exclusions that had been outlined in the appendix to the previous regulations. See 34 C.F.R. Pt. 222, Subpt. K, App. Those regulations have remained materially unchanged since they were promulgated.

2. a. For Fiscal Year 2000, pursuant to 20 U.S.C. 7709(b) and the associated regulations, the Assistant Secretary for Elementary and Secondary Education certified that New Mexico operated an equalized funding program. Pet. App. 41a-42a, 215a-221a. New Mexico, by mandate of state law, seeks to equalize education fund ing among its 89 LEAs. See N.M. Stat. Ann. §§ 22-8-1 et seq. (LexisNexis 1978); Pet. App. 196a-198a. New Mexico allocates its annual appropriation of state educa tion funds among LEAs in a manner generally designed to result in each LEA's having roughly the same overall amount of per-pupil funding, with adjustments made for LEAs that have special funding demands. See Pet. App. 197a-198a, 218a.2

In determining that New Mexico satisfied the 25% disparity standard under the Impact Aid law, the Assis tant Secretary explained that the per-pupil revenues for the LEA at the 95th percentile of all pupils "was $3,259.00 (Penasco) and the per-pupil revenue at the 5th percentile was [$]2,848.00 (Hobbs)." Pet. App. 220a; see id. at 210a-213a. The disparity between those two LEAs is 14.43%, within the 25% threshold. Id. at 220a.

b. Petitioners, two New Mexico LEAs that receive substantial Impact Aid funds, challenged the Assistant Secretary's certification.3 Petitioners argued that New Mexico failed to qualify as equalized under the Impact Aid law, and that the State therefore was barred from taking into account petitioners' Impact Aid funding when allocating state funding among the State's LEAs. Petitioners argued, in particular, that if the 95th and 5th percentile of LEAs-rather than the LEAs serving the 95th and 5th percentile of pupils-were considered, New Mexico's system would not qualify as equalized. An administrative law judge (ALJ) sustained the Assistant Secretary's certification. Pet. App. 43a-58a.

Petitioners appealed the ALJ's decision to the Secre tary, who affirmed the ALJ's decision. Pet. App. 34a- 40a. The Secretary explained that, "[a]lthough the im pact aid statute sets forth the parameters for calculating state public education expenditures or revenues under the disparity test, the statute does not contain a specific implementation of the disparity test." Id. at 37a. Con gress instead had "left that gap to be filled by regula tion, which has been duly promulgated at an appendix to Subpart K of 34 CFR Part 222." Ibid. The Secretary concluded that there "is nothing within the text of the statute that precludes [the regulatory] interpretation or requires another result." Id. at 39a.

3. Petitioners sought judicial review of the Secre tary's determination in the court of appeals under 20 U.S.C. 7711(b). Petitioners contended that the Secre tary's methodology for determining which LEAs to ex clude under the 95th and 5th percentile exclusions con flicts with the terms of the statute, 20 U.S.C. 7709(b)(2)(B)(i).

According to petitioners' argument, the statute pre cludes the Secretary from considering the number of pupils in each LEA when applying the 95th and 5th per centile exclusions. Instead, petitioners asserted, the statute compels the Secretary to eliminate 5% of the LEAs at each end of the spectrum of per-pupil expendi tures or revenues, regardless of the number of pupils served by those LEAs. See Pet. App. 15a. That ap proach would call for eliminating five LEAs (or 5.6% of New Mexico's 89 LEAs) from each end of the spectrum. See id. at 15a n.7. Excluding five LEAs from each end of the range-instead of excluding LEAs until 5% of the pupils are accounted for, as the Secretary's methodol ogy prescribes-would result in excluding LEAs that account for only 0.6% of the State's students at the bot tom of the range and 1.2% of the State's students at the top of the range. See id. at 210a-213a; J.A. 89-92. In other words, rather than measuring equalization based on the LEAs serving 90% of the pupils in the State, peti tioners' approach would measure it based on 98.2% of all students. After excluding five LEAs from each end of the spectrum, the disparity in per-pupil revenues be tween the highest and lowest LEAs would be 26.93%, in excess of the 25% standard. Pet. App. 15a n.7.4

a. A panel of the court of appeals affirmed the Secre tary's decision, rejecting petitioners' contention that the statute precludes the Secretary's methodology. Pet. App. 3a-33a. The panel reasoned that the "statute's am biguity, coupled with the gap left by Congress regarding the specific means by which to implement the disparity test," required that deference be accorded the Secre tary's determination under Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837 (1984). Pet. App. 16a-17a.

The panel concluded that the Secretary's interpreta tion is reasonable, explaining: "Having already ranked the LEAs in descending order by their per-pupil expen ditures, it makes sense that the cut-off points for per centiles would also be based on total student enroll ment." Pet. App. 17a. The panel observed that the Sec retary's "approach supports the basic purpose of the percentile exclusion because it eliminates in a fair and effective manner any unusual or noncharacteristic per- pupil revenues or expenditures that may appear at the extremes of the range of LEAs in the state." Ibid. The panel explained that petitioners' approach, by contrast, "would not further the goal of eliminating [the] unusual distribution of per-pupil expenditures in New Mexico," which "has predominantly small LEAs, several of which rank near the top of per-pupil expenditures." Ibid.

Judge O'Brien dissented from the panel's disposition. Pet. App. 22a-23a. In his view, the statute unambigu ously required a version of petitioners' methodology, under which the Secretary, as petitioners contended, was compelled to eliminate five LEAs from each end of the spectrum of per-pupil revenues. See id. at 24a.5

b. The court of appeals granted rehearing en banc and vacated the panel's opinion. The en banc court is sued a per curiam opinion affirming the Secretary's de cision by an equally divided court, but containing no fur ther explanation or reasoning. Pet. App. 1a-2a.

SUMMARY OF ARGUMENT

The principal issue in this case is whether the Secre tary's methodology for implementing the Impact Aid statute's equalization test is foreclosed by the statutory terms. The plain language of the statute readily encom passes the Secretary's methodology, and related statu tory provisions support, and even explicitly endorse, that methodology.

A State funding program qualifies as equalized under the Impact Aid statute if the disparity in per-pupil reve nues between the highest and lowest ranked LEA is less than 25%. If the statute focused only on the highest and lowest ranked LEAs, the issue in this case would not even arise. The statute further directs, however, that, in making that comparison, the Secretary must "disre gard [LEAs] with per-pupil * * * revenues above the 95th percentile and below the 5th percentiles of such * * * revenues in the State." 20 U.S.C. 7709(b)(2)(B)(i) (emphasis added). The Secretary determines the 95th and 5th percentiles of "such" per-pupil "revenues in the State" by accounting for the number of pupils in the LEAs and identifying the per-pupil revenue figures as sociated with the 95th and 5th percentiles of the total pupil population in the State, and then excluding the LEAs with per-pupil revenues above the 95th percentile and below the 5th percentile.

Nothing in the terms of Section 7709(b)(2)(B)(i) com pels the Secretary to ignore the number of pupils served by an LEA when applying the 95th and 5th percentile exclusions. Contrary to petitioner's argument, the stat ute does not speak in terms of identifying the 95th and 5th percentile "of LEAs." Rather, the statute calls for determining the 95th and 5th percentile "of [per-pupil] * * * revenues in the State." 20 U.S.C. 7709(b)(2)(B)(i). That language does not address the specific methodol ogy for assembling the field of per-pupil revenue figures against which to identify the 95th and 5th percentile "of per-pupil revenues in the State." In particular, the lan guage does not compel the Secretary to assign equal weight to each LEA's per-pupil revenue figure, regard less of the number of pupils served by an LEA. The statute therefore does not preclude the Secretary from applying the 95th and 5th percentile exclusions to focus on 90% of the pupils, as opposed to 90% of the LEAs.

The Secretary's emphasis on the number of pu pils-rather than solely on the number of LEAs-is sup ported by the statutory focus on "per-pupil * * * reve nues in the State" as a whole. 20 U.S.C. 7709(b)(2)(B)(i) (emphasis added). If a State were to divide an existing LEA into two LEAs for purely administrative purposes, with no effect on overall student population or the reve nues allocated to those students, the administrative divi sion would have no substantive effect on the amount or distribution of per-pupil revenues "in the State" as a whole. Yet petitioners' approach, because it focuses on the number of LEAs and assigns equal weight to every LEA, would double the weight assigned to the newly- divided LEA merely because it is now treated for admin istrative purposes as two LEAs rather than one.

In addition, petitioners' interpretation of Section 7709(b)(2)(B)(i) conflicts with Congress's approach in other provisions of the Impact Aid statute and in provi sions governing related programs. The Impact Aid stat ute requires the Secretary to allocate Impact Aid funds to LEAs based on the number of pupils served by an LEA. In that light, Congress presumably did not intend for the Secretary to ignore the number of pupils served by an LEA when determining whether funding is equal ized among LEAs.

Congress made that clear in provisions governing the Education Finance Incentive Grant Program (EFIG), which was established in the same Act of Congress that enacted Section 7709(b)(2)(B)(i). Funding under EFIG is designed to promote equitable education financing by States, and is based in part on the extent of variation in per-pupil expenditures among a State's LEAs. Con gress explicitly required the Secretary, when applying EFIG's formula for evaluating the degree of disparity in per-pupil revenues among a State's LEAs, to measure the variation based on the number of pupils served by an LEA. Congress should not be considered to have simul taneously barred the Secretary from conducting exactly that sort of pupil-based analysis when measuring the degree of variation among LEAs' per-pupil revenues under the Impact Aid statute.

EFIG also supports the Secretary's methodology in the Impact Aid regulations in an even more direct and explicit manner. Congress provided in EFIG that, if a State meets the disparity standard set forth in the Secre tary's Impact Aid regulations-i.e., the very methodol ogy at issue in this case-the State would automatically receive a favorable rating for purposes of EFIG's mea sure of equity in per-pupil expenditures among the State's LEAs. Congress would not have explicitly incor porated in EFIG the methodology set forth in the Secre tary's Impact Aid regulations if Congress, in the same Act, had intended to foreclose the Secretary from using that methodology in the Impact Aid program itself. In light of Congress's explicit endorsement of the Secre tary's methodology, there is no merit to petitioners' ar gument that the statutory history demonstrates an in tent by Congress to reject that approach.

Petitioners' interpretation that the Secretary is barred from considering the number of pupils served by an LEA also is inconsistent with both Congress's gen eral objectives in the equalization inquiry and the spe cific purpose of the 95th and 5th percentile exclusions. Barring the Secretary from taking into account an LEA's pupil population when applying those percentile exclusions could distort the analysis, either by failing to exclude small LEAs with anomalous characteristics that do not materially affect the State's distribution of per- pupil revenues, or by failing to include large LEAs de spite their significant effect on the State's per-pupil rev enues. The uniform view of practitioners in the field of education finance thus is that a disparity test like the one in the Impact Aid statute must take into account the number of pupils served by an LEA. When considered in light of Congress's objectives in the equalization in quiry, accordingly, Section 7709(b)(2)(B)(i) should not-and need not-be read in the manner pressed by petitioners.

In addition, the Secretary's methodology applies a consistent approach across States without regard to idiosyncracies in the size or makeup of LEAs. In any State, the Secretary's methodology measures equaliza- tion by reference to 90% of the students in the State. Petitioners' approach, by contrast, would measure equalization based on 98.2% of all pupils in New Mexico (where the outlier LEAs are relatively small), but would measure it based on less than 90% of pupils in another State where outlier LEAs are relatively large. The Secretary's uniform approach thus makes sense in adminis tering a national program.

Because the statute does not foreclose the Secre tary's methodology for purposes of the first step of the Chevron inquiry, the question under the second step is whether that methodology is reasonable. The Secre tary's methodology readily satisfies that reasonableness standard. There is no merit to petitioners' suggestion that the Secretary's methodology fails to qualify for def erence under Chevron in the first place. Congress di rected the Secretary to make a determination whether a State has in effect a program that equalizes education expenditures, and Congress called for the Secretary to certify a State program that meets the statutory test. Congress plainly gave the Secretary authority to speak with binding force in carrying out those statutory re sponsibilities.

ARGUMENT

THE SECRETARY'S METHODOLOGY FOR DETERMINING WHETHER A STATE OPERATES AN EQUALIZED EDUCA TION FUNDING PROGRAM IS CONSISTENT WITH THE TERMS OF THE STATUTE AND, INDEED, ADVANCES CON GRESS'S OBJECTIVES MORE EFFECTIVELY THAN PETI TIONERS' FLAWED APPROACH

The validity of the Secretary's methodology for de termining whether a State operates an equalized educa tion funding program under 20 U.S.C. 7709 is governed by the two-step framework prescribed by Chevron U.S.A. Inc. v. NRDC, 467 U.S. 837, 842-843 (1984). Peti tioners err in contending that the Secretary's approach is foreclosed at the first stage of the Chevron inquiry. The statute does not unambiguously-and counterintuitively-compel the Secretary to ignore the number of pupils served by a State's LEAs when deter mining whether the State operates an equalized funding program. Rather, the Secretary's decision to consider the number of pupils served by the State's LEAs in de termining which LEAs to exclude from the analysis is fully consistent with the statutory text, is supported by-indeed, explicitly endorsed by-related statutory provisions, and is substantially more effective in advanc ing the statutory objectives than petitioners' flawed ap proach. This Court therefore should sustain the Secre tary's construction.

A. Congress Has Supported, Rather Than Foreclosed, The Secretary's Methodology

The threshold question under Chevron is whether "Congress has directly spoken to the precise question at issue," or whether the statute instead "is silent or am biguous with respect to the specific issue." Chevron, 467 U.S. at 842-843. The question thus is whether "the stat ute unambiguously forecloses the agency's interpreta tion." National Cable & Telecomms. Ass'n v. Brand X Internet Servs., 125 S. Ct. 2688, 2700 (2005). Here, the statutory terms readily encompass the Secretary's methodology.

1. The Secretary's methodology is consistent with the plain language of the statute

The "precise question at issue" for purposes of Chev ron step one, 467 U.S. at 842, is whether, when applying the 95th and 5th percentile exclusions set forth in the statute, 20 U.S.C. 7709(b)(2)(B)(i), the Secretary is re quired to eliminate 5% of the LEAs from each end of the spectrum of LEAs as ranked by per-pupil revenues, or instead may eliminate the outlying five percentiles of pupils as arrayed by per-pupil revenues. The terms of the statute do not unambiguously address that question, much less "unambiguously foreclose" the Secretary's approach. Brand X, 125 S. Ct. at 2700.

a. The statute deems a State to have "in effect a pro gram of State aid that equalizes expenditures" if "the amount of * * * per-pupil revenues available to [the LEA] in the State with the highest such * * * reve nues did not exceed the amount" of "per-pupil revenues available to [the LEA] with the lowest such * * * reve nues by more than 25 percent." 20 U.S.C. 7709(b)(1) and (2)(A).6 If that provision-which appears to focus only on the top and bottom LEAs in the State-stood alone, the issue in this case would not arise. But the statute further provides, in the critical language at issue here, that when applying that 25% disparity standard, the Secretary must "disregard [LEAs] with per-pupil * * * revenues above the 95th percentile or below the 5th per centile of such [i.e., per-pupil] revenues in the State." 20 U.S.C. 7709(b)(2)(B)(i). The pivotal question in this case is how to identify "the 95th percentile" and "the 5th percentile of * * * [per-pupil] revenues in the State." Ibid. (emphasis added).

Petitioners argue that the statute unambiguously requires the Secretary to exclude "the LEAs which fall above the 95th and below the 5th percentiles of LEAs," regardless of the number of pupils served by the ex cluded LEAs. Pet. Br. 22 (emphasis added). The stat ute, however, nowhere speaks in terms of determining the 95th and 5th percentiles "of LEAs." Rather, the statute calls for identifying the 95th and 5th percentiles "of * * * [per-pupil] revenues in the State." 20 U.S.C. 7709(b)(2)(B)(i) (emphasis added). Inasmuch as the statute focuses the analysis on "per-pupil" revenues, there is no basis for reading the text to preclude consid eration of the number of "pupils" in an LEA when apply ing the 95th and 5th percentile exclusions. And nothing in the statute directs that determining the 95th and 5th percentiles "of [per-pupil] revenues in the State" must focus on the universe of LEAs in the State (without re gard to the number of pupils), instead of focusing on the universe of pupils in the State. In other words, the 95th and 5th percentile LEAs are not the same as the LEAs with the 95th and 5th percentile pupils as arrayed by per-pupil revenues, and the statute does not unambigu ously focus on the former.

In particular, the statute does not address whether, in assembling the field of per-pupil revenue figures against which to identify the 95th and 5th percentiles "of such * * * revenues in the State," 20 U.S.C. 7709(b)(2)(B)(i), the Secretary is required to compose an array consisting of one per-pupil revenue for each LEA (as petitioners evidently assume), or instead may com pose an array consisting of one per-pupil revenue for each pupil in the State (as the Secretary's methodology effectively does). Neither of those approaches is unam biguously compelled (or foreclosed) by the text of 20 U.S.C. 7709(b)(2)(B)(i). That text simply requires, with out further elaboration, determining the 95th and 5th percentiles "of [per-pupil] revenues in the State."

b. The difference between the two approaches, as applied to this case, is as follows. Both approaches be gin by ranking the revenues of the State's LEAs, on a per-pupil basis, from lowest to highest. In this case, that list would consist of 89 figures, with one entry for each LEA. Under petitioners' approach, the next step would entail identifying the per-pupil revenue associated with the LEAs that are 5% from each end. Petitioners, that is, attempt to identify the per-pupil revenue associ ated with the 95th and 5th percentiles of LEAs among the 89 LEAs along the spectrum. Petitioners would then assess whether the disparity in the per-pupil reve nues of those two LEAs exceeds 25%.

The Secretary, by contrast, weights each LEA's per- pupil revenue figure by the number of pupils in that LEA. Under that approach, the list of per-pupil reve nues consists not merely of one entry for each of the 89 LEAs (as under petitioners' approach), but instead ef fectively consists of one entry for each pupil in each LEA. The Secretary thus effectively arrays not just the 89 LEAs but the entire population of students in the State. The Secretary next eliminates 5% of those en tries-or 5% of the overall pupil population-from each end of the spectrum. The remaining entry at the high and low end is the per-pupil revenue for the LEA that serves, respectively, the 95th and 5th percentile of pu pils along the spectrum. The Secretary compares the per-pupil revenues of those two LEAs to determine whether it exceeds 25%.7

The Secretary's methodology, no less than petition ers' approach, can readily be described as identifying the 95th and 5th percentiles "of [per-pupil] revenues in the State." 20 U.S.C. 7709(b)(2)(B)(i). Nothing in those terms directs that each and every LEA's per-pupil reve nue figure must be assigned an equal weight in the array of per-pupil revenues, regardless of the relative number of pupils served by an LEA. The statute thus does not foreclose the Secretary's approach of weighting a partic ular LEA's per-pupil revenues by the number of pupils it serves, so as to reflect more accurately the relative contribution of that LEA's revenues to the overall reve nues in the State. See Kern Alexander & Richard G. Salmon, Public School Finance 233 (1995) (explaining that it is "inappropriate[]," when assessing the extent to which education funding is equitably distributed, to "use local school districts as the unit of analysis, thus disre garding the differences in numbers of pupils served by local school districts," and that the analysis instead should "weight the school district proportionally to the number of pupils served").

In New Mexico, for instance, one LEA (Mosquera) serves 57 pupils, and another LEA (Albuquerque) serves over 83,000 pupils. Pet. App. 210a-211a. Petition ers' interpretation of the statute would give the same weight in the analysis to those two LEAs' per-pupil rev enue figures, even though one LEA serves roughly 1500 pupils for every one pupil served by the other. The Sec retary's methodology, by contrast, treats those LEAs not as functionally indistinguishable units but as repre senting distinct populations of pupils. That methodol ogy therefore determines the 95th and 5th percentiles "of * * * [per-pupil] revenues in the State," 20 U.S.C. 7709(b)(2)(B)(i), by reference to which two LEAs' per- pupil revenue figures represent the 95th and 5th percen tiles of pupils in the State.

c. The textual focus on per-pupil revenues "in the State" as a whole, 20 U.S.C. 7709(b)(2)(B)(i), reinforces the conclusion that the statute does not compel the Sec retary to apply the percentile exclusions without regard to the number of pupils served by each of the State's constituent LEAs. That is because that interpretation would attach dispositive significance to adjustments in the number of LEAs even when those adjustments could have no actual effect on the amount or distribution of per-pupil revenues "in the State" as a whole.

A State, for instance, might elect for purely administrative purposes to divide a large LEA into two LEAs, with no change in the total population of pupils served by the newly-divided LEA or in the funds allocated to that fixed population of pupils. The administrative divi sion of the LEA therefore would have no substantive effect on the amount or distribution of per-pupil reve nues "in the State." Petitioners' methodology, however, by focusing exclusively on the number of LEAs and at taching equivalent weight to each LEA, would double the weight given to the per-pupil revenues of the newly- divided LEA merely because it is now regarded as two LEAs rather than one. That purely administrative divi sion thus could be determinative of whether the State is considered to have an equalized distribution of revenues, despite the absence of any substantive consequences for "per-pupil * * * revenues in the State." 20 U.S.C. 7709(b)(2)(B)(i). Congress should not be assumed to have intended-let alone unambiguously intended-that sort of anomalous outcome. See United States v. X- Citement Video, Inc., 513 U.S. 64, 69 (1994) ("Some ap plications of respondents' position would produce results that were not merely odd, but positively absurd * * * . We do not assume that Congress, in passing laws, in tended such results.").

Moreover, the relevant provision directs the Secre tary to ignore certain outlying LEAs in assessing whether per-pupil revenues in the State as a whole are equalized. In such an analysis, it makes sense to focus, as the Secretary's approach does, on the bulk of pupils in the middle and exclude 5% of the students at either extreme. By contrast, it would make little sense to ex clude 5% of the LEAs, without regard to whether that represents only a small percentage of pupils (as in New Mexico because the outliers are small) or a significant percentage of the State's pupils (when the outliers are large). And it is possible that LEAs on one end may be large and LEAs on the other end may be small, which would further skew the analysis. The Secretary's ap proach ensures that, in every State, the degree of equal ization of per-pupil revenues is measured by reference to the revenues associated with the 90% of pupils in the middle of the array.

d. Petitioners err in focusing their analysis of the text (Pet. Br. 16-18) on whether the term "percentile" is ambiguous. For the reasons explained, the relevant am biguity does not concern the meaning of the term "per centile," but instead concerns the precise methodology for assembling the field "of [per-pupil] revenues in the State" against which to apply the 95th and 5th percentile exclusions. 20 U.S.C. 7709(b)(2)(B)(i).8 Petitioners' ap- proach calls for identifying the 95th and 5th percentile of LEAs in a field of 89 observations consisting of one per-pupil revenue figure for each LEA. The Secretary's methodology weights each LEA's per-pupil revenue fig ure by the LEA's pupil population and assembles an array of observations that includes every pupil in the State. The terms of Section 7709(b)(2)(B)(i) do not fore close the latter approach any more than they foreclose the former one. See Graham County Soil & Water Con servation Dist. v. United States, 545 U.S. 409, 419 n.2 (2005) (statute "is ambiguous because its text, literally read, admits of two possible interpretations").9

2. Petitioner's interpretation of the statute conflicts with Congress's approach in related provisions

This Court has instructed that, "[i]n determining whether Congress has specifically addressed the ques tion at issue, a reviewing court should not confine itself to examining a particular statutory provision in isola tion." FDA v. Brown & Wiiliamson Tobacco Corp., 529 U.S. 120, 132 (2000). Here, petitioners' view that the statute forecloses the Secretary from considering the number of pupils served by an LEA is irreconcilable with Congress's approach in related statutory provi sions. See Household Credit Servs., Inc. v. Pfennig, 541 U.S. 232, 241 (2004) (conducting "examination of [the statute's] related provisions" as part of Chevron step- one inquiry).

a. The Impact Aid statute prescribes that, in estab lishing the amount of Impact Aid funds to be provided to any LEA, the Secretary is required to "determine the number of children who were in average daily atten dance in the schools of such agency," and then to deter mine the number of those pupils who have the requisite federal connections. 20 U.S.C. 7703(a)(1). The statute therefore does not treat each recipient LEA as an equiv alent entity, but instead bases an LEA's level of Impact Aid funding on the number of pupils it serves. That un derstanding that LEAs are not interchangeable units but instead serve distinct populations of pupils neces sarily informs the proper interpretation of the Impact Aid statute's equalization test. Congress did not require the Secretary to allocate Impact Aid funds to LEAs based on the number of pupils they serve, while simulta neously requiring the Secretary to assess whether fund ing is equalized among LEAs with no consideration of the number of pupils they serve.

In describing the operation of the 25% disparity test, moreover, the Impact Aid statute not only requires the Secretary to "disregard [LEAs] with per-pupil expendi tures or revenues above the 95th percentile or below the 5th percentile of such expenditures or revenues in the State," 20 U.S.C. 7709(b)(2)(B)(i), but it also requires the Secretary to "take into account the extent to which a program of State aid reflects the additional cost of providing free public education * * * to particular types of students, such as children with disabilities," 20 U.S.C. 7709(b)(2)(B)(ii) (emphasis added). Congress's recognition that an LEA's per-pupil revenue amount alone might fail to account for the costs of educating "particular types of students" reinforces the conclusion that the relative significance of the per-pupil revenues of LEAs can only be considered in the context of the individual pupils served by those LEAs. That context is informed by variations in the number of pupils served by each LEA no less than variations in the "particular types of students" served by each LEA.

b. Any doubt about the permissibility of the Secre tary's interpretation, however, is removed by related provisions enacted in 1994 in the same Act of Congress that enacted the Impact Aid provisions at issue here. In that Act, Congress also established a new Education Finance Incentive Grant Program (EFIG), now codified at 20 U.S.C. 6337 (Supp. III 2003). See Improving Amer ica's Schools Act of 1994, Pub. L. No. 103-382, Tit. I, § 101, 108 Stat. 3575. That program is designed to pro mote equitable education funding by States. Grants to States under EFIG are based on an "equity factor," which-like the equalization test in the Impact Aid statute-aims to measure the degree of variation in per- pupil expenditures among a State's LEAs. 20 U.S.C. 6337(b)(3) (Supp. III 2003).10

In assessing the disparity in per-pupil expenditures among LEAs for purposes of EFIG's equity factor, Con gress required calculation of a "coefficient of variation for the per-pupil expenditures of local educational agen cies," 20 U.S.C. 6337(b)(3)(A)(ii)(I) (Supp. III 2003), which, like the 25% disparity test in the Impact Aid stat ute, is one method of measuring the degree of variation among per-pupil expenditures in a State. See Public School Finance 236 (describing both approaches). Of particular relevance here, Congress directed that, in applying that measure, "the Secretary shall weigh the variation between per-pupil expenditures in each [LEA] * * * according to the number of pupils served by the [LEA]." 20 U.S.C. 6337(b)(3)(A)(ii)(II) (Supp. III 2003) (emphasis added). Given that Congress compelled the Secretary to consider an LEA's number of pupils when assessing the disparity among per-pupil expenditures for purposes of EFIG, it would require much clearer language than the terms of Section 7709(b)(2)(B)(i) to conclude that Congress compelled the Secretary to ig nore an LEA's number of pupils when assessing the dis parity among per-pupil expenditures for purposes of the Impact Aid program. That is particularly the case be cause the two sets of provisions were enacted by the same Act of Congress and serve the same function.

What is more, Congress went further in the provi sions of EFIG and explicitly endorsed the Secretary's methodology for applying the disparity test under the Impact Aid program. Congress provided in EFIG that a State "that meets the disparity standard described in section 222.63 of title 34, Code of Federal Regulations (as such section was in effect on the day preceding Octo ber 10, 1994)"-i.e., a State that qualifies as equalized under the Secretary's methodology in the Impact Aid regulations-would automatically be granted a favorable equity rating for purposes of EFIG. 20 U.S.C. 6336(b)(3)(B) (1994).11 Congress cannot be considered to have explicitly endorsed the Secretary's Impact Aid methodology by essentially incorporating it into the pro visions of EFIG, but to have simultaneously-in the same enactment-implicitly prohibited that same meth odology under the Impact Aid program itself.

Finally, after the Secretary promulgated new Impact Aid regulations in the wake of the 1994 statute, in which he re-issued the methodology for implementing the dis parity standard but under a different code section (i.e., 34 C.F.R. 222.162), Congress once again demonstrated its endorsement of the Secretary's methodology by en acting an amendment to make a corresponding adjust ment in the terms of EFIG. See 20 U.S.C. 6337(b)(3)(B) (Supp. III 2003) (referring to "the disparity standard described in section 222.162 of title 34, Code of Federal Regulations") (enacted by the No Child Left Behind Act of 2001, Pub. L. No. 107-110, § 101, 115 Stat. 1527).

3. Petitioners' interpretation of the statute is inconsis tent with Congress's objectives

As the Court has explained, "[i]n determining the meaning of the statute, we look not only to the particular statutory language, but to the design of the statute as a whole and to its object and policy." Crandon v. United States, 494 U.S. 152, 158 (1990); see Robinson v. Shell Oil Co., 519 U.S. 337, 341 (1997) (explaining that the "ambiguity of statutory language is determined" not only "by reference to the language itself," but also by reference to "the specific context in which that language is used, and the broader context of the statute as a whole"). Here, the broader statutory objective is to de termine whether a State funding program equalizes the distribution of per-pupil revenues across the State, and the specific purpose of the 95th and 5th percentile exclu sions is to eliminate anomalous characteristics of outly ing observations at the extremes of the range that could distort the analysis.

a. Petitioners' interpretation that the Secretary is barred from considering the number of pupils served by each LEA stands at cross purposes with the object of the percentile exclusions. For precisely that reason, the Secretary rejected a methodology that would exclude a specific number (or percentile) of LEAs without consid ering the number of students served by those LEAs:

[B]asing an exclusion on numbers of districts would act to apply the disparity standard in an unfair and inconsistent manner among the States. The purpose of the exclusion is to eliminate those anomalous char acteristics of a distribution of expenditures. In States with a small number of large districts, an ex clusion based on percentage of school districts might exclude from the measure of disparity a substantial percentage of the pupil population in those States. Conversely, in States with large numbers of small districts, such an approach might exclude only an insignificant fraction of the pupil population and would not exclude anomalous characteristics.

41 Fed. Reg. 26,324 (1976).

For instance, if a State has a number of small LEAs ranking at the top and bottom of the range of per-pupil revenues-as is the case in New Mexico, see pp. 11-12, supra-eliminating only 5% of the LEAs could exagger ate the degree of disparity actually experienced by the lion's share of pupils in the State. Congress's objective was to eliminate outlying observations and focus the analysis on the degree of disparity experienced by the bulk (i.e., the middle 90%) of students in the State.12

As an example, if the 5% of LEAs that are excluded under petitioner's approach together with the LEAs at each end of the remaining spectrum account for less than one percent of the overall pupil population, the State might fail the 25% disparity test, and thus be deemed non-equalized, even though it has achieved equalized funding with respect to 99% of its pupils and 99% of its overall per-pupil revenues. And in such a sit uation, the small districts at the ends of the remaining range that would be responsible for the determination that the State fails the 25% disparity test are especially likely to display the sorts of anomalous characteristics that should exclude them from the analysis. See Public School Finance 233 ("states occasionally experience exceptionally high per-pupil costs to provide educational services for limited numbers of pupils").

Conversely, when a State's highest (or lowest) LEAs in terms of per-pupil revenues serve disproportionately large numbers of pupils, excluding 5% of the LEAs from each end of the range-without regard to the number of pupils excluded thereby-could give a false impression that the State operates an effective equalization pro gram. For instance, if the LEAs comprising the highest and lowest 5% of the range of per-pupil revenues consist of large LEAs accounting for a majority of the pupils in the State, the State could be deemed equalized based on the remaining LEAs even if there were substantial dis parities in per-pupil revenues for a majority of the State's pupils.13

Moreover, as the Secretary emphasized, differences in the relative sizes of outlying LEAs can vary across States, such that the disparity standard, under petition ers' approach, would be applied "in an unfair and incon sistent manner among the States." 41 Fed. Reg. at 26,324. In a national program, it only makes sense to adopt a standard that in every State focuses on the ques tion whether the bulk of students-the same 90% of stu dents in the middle-suffer from disparities in education financing. For those reasons, the ineffectiveness of peti tioners' approach in advancing the statutory objectives counsels strongly against reading 20 U.S.C. 7709(b)(2)(B)(i) to compel that result.

b. The anomalies produced by petitioners' interpre tation also explain the uniform view of practitioners in the field of education finance that an equalization test like that prescribed by Section 7709(b)(2)(B) must con sider the number of pupils associated with a particular school district's per-pupil revenues, rather than simply give equal weight to every school district's per-pupil revenue amount. In the administrative proceedings in this case, a leading authority in the field accordingly explained: "It is the practice in the field in using tests of this kind to exclude the segments of the student popu lation at the margins of the ranking so as to avoid having their per-pupil input (i.e., expenditure or revenue) falsely overstate disparities." J.A. 5 para. 6. As a re sult, the Secretary's approach is "methodologically sound" in "excluding LEAs in the ranking that fell above or below the 95th and 5th percentiles of the total num ber of students in the State." J.A. 4-6 para. 7. By con trast, that leading authority explained, petitioners' ap proach of excluding "the top and bottom 5 percentiles of expenditures or revenues per pupil in lieu of pupils" is "inappropriate," because that "method would not pre vent the problem of false disparities" and would "render the test ineffective to gauge whether a State's program is achieving equity." J.A. 6-7 para. 10.

Petitioners do not dispute that the Secretary's meth odology reflects the accepted practice in the field.14 Congress should not be assumed to have compelled the Secretary to abandon a methodology universally en dorsed by experts in the field and to adopt an approach uniformly regarded by them as deficient.15

4. The statutory history demonstrates that Congress endorsed the Secretary's methodology

Petitioners argue (Pet. Br. 6-15, 24-25) that the "leg islative and regulatory history" of Section 7709(b)(2)(B)(i) indicates an unambiguous intention of Congress to preclude the Secretary from considering an LEA's number of pupils when applying the 95th and 5th percentile exclusions. The terms of that provision do not specifically resolve the matter, however, and related statutory provisions affirmatively support the Secretary's approach, which should be decisive for purposes of taking the analysis past Chevron's first step. Peti tioners, in any event, draw precisely the wrong infer ences from that history.

Petitioners assume that, when Congress enacted Sec tion 7709(b)(2)(B)(i) in 1994, Congress rejected the methodology that had been set forth in the Secretary's regulations since 1976. Petitioners' argument is grounded in the fact that the language of the regulations made explicit that the application of the 95th and 5th percentile exclusions turned on the number of pupils, whereas the language of the statute does not speak di rectly to that issue. Compare 20 U.S.C. 7709(b)(2)(B)(i) (referring to 95th and 5th percentiles "of such [per-pu pil] expenditures or revenues in the State"), with 34 C.F.R. Pt. 222, Subpt. K, App. (1993) (referring to "the 95th and 5th percentiles of the total number of pupils in attendance" in the State's LEAs). The ambiguity in Section 7709(b)(2)(B)(i), however, could hardly be viewed as an unambiguous rejection of the Secretary's longstanding methodology. Cf. Green v. Bock Laundry Mach. Co., 490 U.S. 504, 521 (1989) ("A party contending that legislative action changed settled law has the bur den of showing that the legislature intended such a change.").

That the provision cannot be read as an implicit re jection of the Secretary's approach is especially clear given that the Secretary himself proposed the relevant language, as part of his proposed legislation to reauthorize the Impact Aid program. See pp. 6-7, su pra. The Secretary of course gave no indication that he viewed his own proposal to constrain his ability to con tinue the agency's consistent approach of almost two decades and to require him instead to use a methodology he had long since rejected as unsound and inequitable. And Congress gave no indication that it viewed the Sec retary's proposal to have that disruptive effect. Cf. Miller v. Youakim, 440 U.S. 125, 144 (1979) ("Adminis trative interpretations are especially persuasive where, as here, the agency participated in developing the provi sion."). The Secretary's actions following the enactment of the language he proposed confirm that no change was intended. The Secretary promulgated new regulations generally implementing the reauthorization legislation, but retained, essentially without change, the previous regulations concerning application of the 25% disparity test and the 95th and 5th percentile exclusions. See 34 C.F.R. 222.162(a); 34 C.F.R. Pt. 222, Subpt. K. App.

Finally, any suggestion that Congress somehow in tended to disapprove of the Secretary's methodology when it enacted the Secretary's own proposal is conclu sively refuted by the fact that, in the same Act in which it enacted the Secretary's proposal, Congress also ex plicitly incorporated and endorsed the Secretary's meth odology in the provisions establishing EFIG. See pp. 29-30, supra. In that light, the legislative and regula tory history relied on by petitioners, far from showing a clear intention to reject the Secretary's methodology, in fact demonstrates an intention to preserve it.

B. The Secretary's Methodology Is Based On A Reasonable Construction Of The Statute

Because the statute, for purposes of the first step of the Chevron framework, contains no unambiguous prohi bition against considering the number of pupils in an LEA when applying the 95th and 5th percentile exclu sions, the question under the second step of Chevron is whether the Secretary's methodology rests on a reason able construction of the statute. See Chevron, 467 U.S. at 844. Petitioners understandably devote the bulk of their attention to the question whether the statute un ambiguously precludes the Secretary's approach. There could be no serious question that, insofar as it is permit ted by the statutory terms, the Secretary's methodology amply qualifies as a "reasonable policy choice for the [agency] to make." Brand X, 125 S. Ct. at 2708 (quoting Chevron, 467 U.S. at 845).

As the Secretary explained when initially establish ing the methodology in 1976, that approach facilitates a sound analysis of whether a State's education funding is equalized by eliminating anomalous characteristics that may lie at the extremes of the range of per-pupil expen ditures. 41 Fed. Reg. at 26,324. The approach is sub stantially superior to petitioners' competing methodol ogy in all the respects previously explained, and it has been applied consistently for a period of three decades. See Smiley v. Citibank (South Dakota), N.A., 517 U.S. 735, 740 (1996) ("[A]gency interpretations that are of long standing come before us with a certain credential of reasonableness, since it is rare that error would long persist."). In addition, the methodology addresses a quintessentially "technical" and "complex" subject mat ter of the kind ordinarily left to the agency's expertise. National Cable & Telecomms. Ass'n v. Gulf Power Co., 534 U.S. 327, 339 (2002). In short, the Secretary's meth odology not only is reasonable, but is far sounder than petitioners' approach as a means to fulfill the Secre tary's responsibility to determine whether a "State has in effect a program of State aid that equalizes expendi tures for free public education among [LEAs] in the State." 20 U.S.C. 7709(b)(1).

C. The Secretary's Regulations Are Entitled To Deference Under The Chevron Framework

While petitioners principally argue that the Secre tary's regulation is inconsistent with the text of the stat ute for purposes of the first step of Chevron, petitioners also suggest that the Secretary's regulations do not war rant consideration under the Chevron framework in the first place. Pet. Br. 37-44. In the court of appeals, how ever, petitioners affirmatively "agree[d] that the analy sis of this case begins with Chevron." C.A. En Banc Pet. 3. Petitioners, in any event, are fundamentally mistaken in contending that Chevron does not apply in this case.

Petitioners appear to argue that Congress did not confer authority on the Secretary to adopt binding rules concerning administration of the Impact Aid statute's equalization test. Insofar as petitioners' contention is that the Secretary's methodology falls outside the scope of her authority because it is foreclosed by the statute, see, e.g., Pet. Br. 41, the argument collapses into petition ers' (erroneous) arguments concerning step one of the Chevron framework. To the extent petitioners suggest more broadly that the Secretary lacks authority to act with binding force in administering the equalization in quiry, petitioners are wrong.

The statute explicitly directs the Secretary to "mak[e] a determination," 20 U.S.C. 7709(b)(2)(B), of whether a "State has in effect a program of State aid that equalizes expenditures," 20 U.S.C. 7709(b)(1), and to "certify the program" if "the Secretary determines that a program of State aid qualifies" as equalized, 20 U.S.C. 7709(c)(3)(A). Those provisions plainly provide for the Secretary to speak with authoritative force in carrying out her statutory responsibilities. See 20 U.S.C. 7709(d)(2) (prohibiting the State from taking Im pact Aid payments into consideration until the Secretary certifies the State's program); see also 20 U.S.C. 1221e-3 ("The Secretary, in order to carry out functions other wise vested in the Secretary by law or by delegation of authority pursuant to law, * * * is authorized to make, promulgate, issue, rescind, and amend rules and regula tions governing the manner of operation of, and govern ing the applicable programs administered by, the De partment."); 20 U.S.C. 3474.

Petitioners fare no better in contending (Pet. Br. 43- 44) that the Secretary's determination should be denied Chevron deference because the regulations embodying the Secretary's methodology, when issued in 1995, were not promulgated pursuant to notice-and-comment proce dures. As petitioners themselves explain (Pet. Br. 13-15, 44), those regulations were essentially a re-issuance of regulations that had initially been promulgated in 1976, and those preexisting regulations were issued through notice-and-comment procedures, see 41 Fed. Reg. at 26,320, 26,329; see also 60 Fed. Reg. 50,778 (1995) (ex plaining that 1995 "regulations do not establish or affect substantive policy"). And even in the normal course, when there is no such preexisting notice-and-comment regulation, the Court has made clear that "deference under Chevron * * * does not necessarily require an agency's exercise of express notice-and-comment rulemaking power." Edelman v. Lynchburg Coll., 535 U.S. 106, 114 (2002).

In addition, even if there were any doubt about the deference owed to the regulations, the administrative action under review in this case is not the issuance of the 1995 regulations as such, but instead is the formal deci sion by the Secretary concluding that New Mexico is entitled to certification of its equalization program un der 20 U.S.C. 7709(c)(3). Pet. App. 34a-40a. The Secre tary authoritatively determined in that decision that the methodology set forth in the regulations is an appropri ate interpretation of the statute and that New Mexico was entitled to certification under the statute. See id. at 38a (concluding that "New Mexico's program [c]omplied with the statutory requirements"); id. at 40a ("[T]hose regulations are consist[e]nt with the statutory provision they implement."). That formal determination by the Secretary is fully entitled to treatment under the Chev ron framework.

Finally, in the circumstances of this case, any ques tion concerning whether the Secretary's methodology is entitled to be judged for reasonableness under Chevron is largely an academic one. Insofar as the statute does not foreclose the Secretary's methodology, her construc tion is not merely a reasonable one, but a plainly supe rior one.

CONCLUSION

The judgment of the court of appeals should be affirmed.

Respectfully submitted.

PAUL D. CLEMENT
Solicitor General

PETER D. KEISLER
Assistant Attorney General

EDWIN S. KNEEDLER
Deputy Solicitor General

SRI SRINIVASAN
Assistant to the Solicitor
General

ROBERT S. GREENSPAN
PETER R. MAIER
Attorneys
Department of Justice

KENT D. TALBERT
General Counsel

STEPHEN H. FREID
MARK SMITH
Attorneys
Department of Education

DECEMBER 2006

1 The Secretary's proposal had retained the exclusion at the top of the range for the 95th percentile, but had not proposed retaining the corresponding exclusion at the bottom of the range. Cf. Richard G. Salmon, The Measurement of Fiscal Equalization Pursuant to Federal Impact Aid, P.L. 81-874, Section 5(d)(2): Recommendations for Improvement, 18 J. Educ. Fin. 18, 30 (1992) (suggesting that "there is no reason to exclude school districts from the analysis that incur low per-pupil costs").

2 Although virtually all States attempt in some measure to amelio rate revenue disparities among school districts, see GAO, State Efforts To Reduce Funding Gaps Between Poor and Wealthy Districts 6-7 (1997), in recent years, only three States-New Mexico, Alaska, and Kansas-have sought and obtained the Secretary's certification that they operate an equalized funding program for purposes of the Impact Aid program. Certain States may choose not to seek certification because the State's LEAs receive modest amounts of Impact Aid funds, such that the State's distribution of state aid is not materially affected by whether it can take into account Impact Aid funding. In addition, the ability of a State to promote equalized funding may be affected by, inter alia, the extent to which funding of LEAs in the State is com prised of state aid as opposed to local revenues, and the degree to which the State elects to target state aid to offset disparities in local revenues. See 1998 GAO Report 9-12. New Mexico's ability to achieve equaliza tion rests in part on the fact that the share of LEA funding that comes from State aid (as opposed to local property tax revenues) is among the highest in the nation. See id. at 14 ("State contributions in the 1991-92 school year ranged from 8 percent of total (state and local) funding in New Hampshire to 85 percent of total funding in New Mexico.").

3 Thirty of New Mexico's 89 LEAs received Impact Aid funds in fis cal year 2000. Pet. App. 234a-235a. The amount of Impact Aid funding received by an individual New Mexico LEA ranged from a low of $106 for one LEA to a high of almost $22.5 million for petitioner Gallup- McKinley County School District No. 1. Ibid. The Impact Aid funding received by petitioners' two LEAs amounts to almost one-half of the total Impact Aid funding received by all of the States' LEAs. See ibid.

4 Petitioners noted that 5% of New Mexico's 89 LEAs would fall somewhere between four and five LEAs: four LEAs would equal 4.5% of the 89 LEAs and five LEAs would equal 5.6% of the LEAs. Petitioners contended that five LEAs should be excluded under their approach. Pet. C.A. En Banc Reply Br. 4. The distinction is immaterial in this case as a practical matter, because exclusion of either four or five LEAs from each end of the spectrum would produce a disparity mea- sure in excess of 25%.

5 Petitioners explained below that Judge O'Brien's approach "followed the same process" as petitioners' approach of eliminating 5% of the LEAs-or five LEAs-from each end of the spectrum. Pet. C.A. En Banc Reply Br. 4; id. at 5 ("Both procedures essentially performed the same calculation."). Judge O'Brien, however, as an intermediate step, used the Microsoft Excel software program to calculate values for the 5th and 95th percentiles of per-pupil revenues, as applied to an array consisting of one per-pupil revenue figure for each of New Mexico's 89 LEAs. See Pet. App. 24a & n.11. He then excluded those LEAs whose per-pupil revenues fell below or above the values for the 5th and 95th percentiles as calculated by the software. The ultimate result was to exclude the same five LEAs from each end of the range as under petitioners' approach of excluding 5% of the LEAs. It appears that Judge O'Brien's approach generally would give rise to the same outcomes as petitioners' shorthand approach of arraying the LEAs and eliminating 5% of the LEAs from each end of the spectrum (but without the intermediate step of calculating a precise value for the 5th and 95th percentiles of per-pupil revenues). In this Court, petitioners continue to describe Judge O'Brien's approach as interchangeable with their own. See Pet. Br. 22-24. We therefore treat the approaches as equivalent, and they are clearly equivalent in the most relevant sense, in that they both look directly to the 95th and 5th percentile of LEAs, without accounting for the number of pupils served by an LEA. See pp. 21-24, infra.

6 The statute allows for making that calculation based on either per- pupil expenditures or per-pupil revenues. See 20 U.S.C. 7709(b)(2)(A) and (B)(i). For ease of reference, and because the calculation for New Mexico is based on per-pupil revenues, we focus on per-pupil revenues rather than per-pupil expenditures.

7 Petitioners suggest (Pet. Br. 19 n.6) that there is a fundamental difference between a "percentile" and a "percentage," and that it is thus inappropriate to equate (i) excluding 5% of the pupils from each end of the spectrum, with (ii) excluding pupils above and below the 95th and 5th percentiles of pupils along the spectrum. There is an obvious relationship between percentiles and percentages, however, such that those two inquiries are functionally similar. See Wilfred J. Dixon & Frank J. Massey, Jr., Introduction to Statistical Analysis 9 (4th ed. 1983) (explaining that "the 10th percentile * * * is defined as the value below which 10 percent of the distribution of values will fall").

8 The point is not that the term percentile is ambiguous, but that it begs the question-percentiles of what? The statute does not, as petitioners would have it, specify percentiles of the LEAs. Rather, it addresses percentiles of per-pupil revenues. In arraying the observa tions of per-pupil revenues, it could be possible to array observations only of LEAs or to array observations of pupils. Because the statute ultimately is concerned about measuring disparity among revenues for pupils across a State and more broadly about educating pupils, it is reasonable for the Secretary to choose the latter approach.

9 When the case was before the court of appeals' panel, petitioner Zuni argued that the Secretary, as an alternative approach to eliminat ing 5% of the LEAs from each end of the spectrum, could calculate 95% of the per-pupil revenues for the LEA with the highest such revenues and eliminate any LEA whose per-pupil revenues exceeded that amount, and also apply a corresponding exclusion at the low end of the spectrum. See Pet. App. 15a. (Petitioner Gallup-McKinley made no challenge to the Secretary's methodology at the panel stage below, but instead made an unrelated argument based on a separate statutory provision. See id. at 19a-22a.) Although petitioners largely avoid reiterating that alternative approach in this Court, at one point petitioners describe that alternative as "precisely what the statute requires." Pet. Br. 26 (discussing method that entails applying 95th and 5th percentile exclusions "to the revenues per membership of the highest and lowest LEAs in the State and then excluding LEAs with per-pupil revenues above or below the product of those calculations"). That methodology does confuse percentiles with percentages, cf. note 7, supra, and gives rise to substantially different results than the alternative of eliminating 5% of the LEAs from each end of the spectrum. See Pet. App. 15a n.7.

10 In addition to the equity factor, the level of funding also turns on an "effort factor," which measures the extent to which a State uses avail able resources to fund education. 20 U.S.C. 6337(b)(2) (Supp. III 2003).

11 The Conference Report specifically observed that the favorable equity rating would apply "[i]f a State meets the expenditure disparity standard under the Impact Aid program regulations (currently Alaska, Kansas and New Mexico)." H.R. Conf. Rep. No. 761, 103d Cong., 2d Sess. 639 (1994) (emphasis added).

12 The number of pupils served by a particular LEA varies widely among and within individual States, and a substantial number of LEAs serve a very small pupil population. In New Mexico itself, for instance, the number of pupils served by any one school district is as small as 57 pupils and as large as more than 83,000 pupils. See Pet. App. 210a- 211a; see also Public School Finance 233 (noting that "Ohio maintains three island districts, [with] each district enrolling no more than six pupils," and observing that Virginia has school districts that serve from 350 to 135,000 pupils each).

13 In New Mexico, roughly 10% of the State's LEAs (or nine of the 89 LEAs)-i.e., Albuquerque, Las Cruces, Gallup-McKinley, Gadsen, Farmington, Roswell, Clovis, Rio Rancho, and Hobbs-account for over 56% of the State's pupils. See Pet. App. 210a-213a; J.A. 89-92.

14 See, e.g., Public School Finance 233 ("The unit of analysis should be based upon pupils and not the local school district," because the focus in "equity analysis * * * should be on pupils served throughout the state and not on the administrative structure that serves only as a vehicle for delivery of educational services. In the case of Virginia, by treating school districts as the unit of analysis, Highland County Public Schools, serving approximately 350 pupils, would exert an identical statistical influence as Fairfax County, serving approximately 135,000 pupils."); Allan R. Odden & Lawrence O. Picus, School Finance: A Policy Perspective 50 (2d ed. 2000) ("A statistical solution is to 'weight' the district or site measure by the number of students," because "[i]f this statistical weighting is not done, each district regardless of size is treated as one observation. Thus, in New York state for example, New York City with a million students and about one-third of all students in the state would affect the statistical findings exactly as much as would a small, rural district with only 100 students. This simply does not make sense."); Robert Berne & Leanna Stiefel, The Measurement of Equity in School Finance: Conceptual, Methodological, and Empiri cal Dimensions 59 (1984) (explaining that "pupil unit of analysis pre- dominates" over "district unit of analysis," and "it seems to us that each pupil should receive equal weight regardless of the size of the district in which she or he is enrolled").

15 Cf. 20 U.S.C. 6336(3)(C) (2000) ("Secretary may revise each State's equity factor as necessary based on the advice of independent education finance scholars").

 

APPENDIX

STATUTORY AND REGULATORY
PROVISIONS INVOLVED

1. 20 U.S.C. 6337 provides, in pertinent part:

§ 6337. Education finance incentive grant program

(a) Grants

From funds appropriated under subsection (f) of this section the Secretary is authorized to make grants to States, from allotments under subsection (b) of this sec tion, to carry out the programs and activities of this part.

(b) Distribution based upon fiscal effort and equity

* * * * *

(3) Equity factor

(A) Determination

(i) In general

Except as provided in subparagraph (B), the Secretary shall determine the equity factor under this section for each State in accordance with clause (ii).

(ii) Computation

(I) In general

For each State, the Secretary shall compute a weighted coefficient of variation for the per- pupil expenditures of local educational agen cies in accordance with subclauses (II), (III), and (IV).

(II) Variation

In computing coefficients of variation, the Secretary shall weigh the variation between per- pupil expenditures in each local educational agency and the average per-pupil expenditures in the State according to the number of pupils served by the local educational agency.

* * * * *

(B) Special rule

The equity factor for a State that meets the disparity standard described in section 222.162 of title 34, Code of Federal Regulations (as such section was in effect on the date preceding Janu ary 8, 2002) or a State with only one local educa tional agency shall be not greater than 0.10.

* * * * *

2. 20 U.S.C. 7709 provides, in pertinent part:

§ 7709. State consideration of payments in providing State aid

(a) General prohibition

Except as provided in subsection (b) of this section, a State may not-

(1) consider payments under this subchapter in determining for any fiscal year-

(A) the eligibility of a local educational agency for State aid for free public education; or

(B) the amount of such aid; or

(2) make such aid available to local educational agencies in a manner that results in less State aid to any local educational agency that is eligible for such payment than such agency would receive if such agency were not so eligible.

(b) State equalization plans

(1) In general

A State may reduce State aid to a local educational agency that receives a payment under section 7702 or 7703(b) of this title (except the amount calculated in ex cess of 1.0 under section 7703(a)(2)(B) of this title and, with respect to a local educational agency that receives a payment under section 7703(b)(2) of this title, the amount in excess of the amount that the agency would receive if the agency were deemed to be an agency eligi ble to receive a payment under section 7703(b)(1) of this title and not section 7703(b)(2) of this title) for any fiscal year if the Secretary determines, and certifies under subsection (c)(3)(A) of this section, that the State has in effect a program of State aid that equalizes expenditures for free public education among local educational agen cies in the State.

(2) Computation

(A) In general

For purposes of paragraph (1), a program of State aid equalizes expenditures among local educational agencies if, in the second fiscal year preceding the fis cal year for which the determination is made, the amount of per-pupil expenditures made by, or per-pu pil revenues available to, the local educational agency in the State with the highest such per-pupil expendi tures or revenues did not exceed the amount of such per-pupil expenditures made by, or per-pupil revenues available to, the local educational agency in the State with the lowest such expenditures or revenues by more than 25 percent.

(B) Other factors

In making a determination under this subsection, the Secretary shall-

(i) disregard local educational agencies with per- pupil expenditures or revenues above the 95th percentile or below the 5th percentile of such expenditures or reve nues in the State; and

(ii) take into account the extent to which a pro gram of State aid reflects the additional cost of pro viding free public education in particular types of lo cal educational agencies, such as those that are geo graphically isolated, or to particular types of stu dents, such as children with disabilities.

* * * * *

(c) Procedures for review of State equalization plans

(1) Written notice

(A) In general

Any State that wishes to consider payments de scribed in subsection (b)(1) of this section in pro viding State aid to local educational agencies shall submit to the Secretary, not later than 120 days before the beginning of the State's fiscal year, a written notice of the State's intention to do so.

(B) Contents

Such notice shall be in the form and contain the information the Secretary requires, including evi dence that the State has notified each local educa tional agency in the State of such State's intention to consider such payments in providing State aid.

(2) Opportunity to present views

Before making a determination under subsection (b) of this section, the Secretary shall afford the State, and local educational agencies in the State, an opportunity to present their views.

(3) Qualification procedures

If the Secretary determines that a program of State aid qualifies under subsection (b) of this sec tion, the Secretary shall-

(A) certify the program and so notify the State; and

(B) afford an opportunity for a hearing, in accor dance with section 7711(a) of this title, to any local educational agency adversely affected by such cer tification.

(4) Nonqualification procedures

If the Secretary determines that a program of State aid does not qualify under subsection (b) of this section, the Secretary shall-

(A) so notify the State; and

(B) afford an opportunity for a hearing, in accor dance with section 7711(a) of this title, to the State, and to any local educational agency adversely af fected by such determination.

(d) Treatment of State aid

(1) In general

* * * * *

(2) Prohibition

A State may not take into consideration payments under this subchapter before such State's program of State aid has been certified by the Secretary un der subsection (c)(3) of this section.

* * * * *

3. 34 C.F.R. 222.162 provides, in pertinent part:

What disparity standard must a State meet in order to be certified and how are disparities in current expendi tures or revenues per pupil measured?

(a) Percentage disparity limitation. The Secretary considers that a State aid program equalizes expendi tures if the disparity in the amount of current expendi tures or revenues per pupil for free public education among LEAs in the State is no more than 25 percent. In determining the disparity percentage, the Secretary disregards LEAs with per pupil expenditures or reve nues above the 95th or below the 5th percentile of those expenditures or revenues in the State. The method for calculating the percentage of disparity in a State is in the appendix to this subpart.

* * * * *

4. 34 C.F.R. Pt. 222, Subpt. K. App. provides, in pertinent part:

APPENDIX TO SUBPART K OF PART 222-DETERM- INATIONS UNDER SECTION 8009 OF THE ACT-METHODS OF CALCULATIONS FOR TREAT MENT OF IMPACT AID PAYMENTS UNDER STATE EQUALIZATION PROGRAMS

The following paragraphs describe the methods for making certain calculations in conjunction with deter minations made under the regulations in this subpart. Except as otherwise provided in the regulations, these methods are the only methods that may be used in mak ing these calculations.

1. Determinations of disparity standard compliance under § 222.162(b)(1).

(a) The determinations of disparity in current expendi tures or revenue per pupil are made by-

(i) Ranking all LEAs having similar grade levels within the State on the basis of current expenditures or revenue per pupil for the second preceding fiscal year before the year of determination;

(ii) Identiyfing those LEAs in each ranking that fall at the 95th and 5th percentiles of the total number of pupils in attendance in the schools of those LEAs; and

(iii) Subtracting the lower current expenditure or revenue per pupil figure from the higher for those agencies identified in paragraph (ii) and dividing the difference by the lower figure.

Example: In State X, after ranking all LEAs orga nized on a grade 9-12 basis in order of the expenditures per pupil for the fiscal year in question, it is ascertained by counting the number of pupils in attendance in those agencies in ascending order of expenditure that the 5th percentile of student population is reached at LEA A with a per pupil expenditure of $820, and that the 95th percentile of student population is reached at LEA B with a per pupil expenditure of $1,000. The percentage disparity between the 95th and 5th percentile LEAs is 22 percent ($1,000-$820 = $180/$820).

* * * * *