Arkansas v. Oklahoma/Opinion of the Court

Justice Stevens delivered the opinion of the Court.

Pursuant to the Clean Water Act, 86 Stat. 816, as amended, 33 U.S.C. §1251 et seq., the Environmental Protection Agency (EPA or Agency) issued a discharge permit to a new point source in Arkansas, about 39 miles upstream from the Oklahoma state line. The question presented in this litigation is whether the EPA's finding that discharges from the new source would not cause a detectable violation of Oklaho- '''[p. 95]''' ma's water quality standards satisfied the EPA's duty to protect the interests of the downstream State. Disagreeing with the Court of Appeals, we hold that the Agency's action was authorized by the statute.

I
In 1985, the city of Fayetteville, Arkansas, applied to the EPA, seeking a permit for the city's new sewage treatment plant under the National Pollution Discharge Elimination System (NPDES). After the appropriate procedures, the EPA, pursuant to §402(a)(1) of the Act, 33 U.S.C. §1342(a)(1), issued a permit authorizing the plant to discharge up to half of its effluent (to a limit of 6.1 million gallons per day) into an unnamed stream in northwestern Arkansas. That flow passes through a series of three creeks for about 17 miles, and then enters the Illinois River at a point 22 miles upstream from the Arkansas-Oklahoma border.

The permit imposed specific limitations on the quantity, content, and character of the discharge and also included a number of special conditions, including a provision that if a study then underway indicated that more stringent limitations were necessary to ensure compliance with Oklahoma's water quality standards, the permit would be modified to incorporate those limits. App. 84.

Respondents challenged this permit before the EPA, alleging, inter alia, that the discharge violated the Oklahoma water quality standards. Those standards provide that "no degradation [of water quality] shall be allowed" in the upper Illinois River, including the portion of the river immediately downstream from the state line.

'''[p. 96]' Following a hearing, the Administrative Law Judge (ALJ) concluded that the Oklahoma standards would not be implicated unless the contested discharge had "something more than a mere de minimis'' impact" on the State's waters. He found that the discharge would not have an "undue impact" on Oklahoma's waters and, accordingly, affirmed the issuance of the permit. App. to Pet. for Cert. in No. 90–1262, pp. 101a–103a (emphasis deleted).

On a petition for review, the EPA's Chief Judicial Officer first ruled that §301(b)(1)(C) of the Clean Water Act "requires an NPDES permit to impose any effluent limitations necessary to comply with applicable state water quality standards." Id., at 116a–117a. He then held that the Act '''[p. 97]''' and EPA regulations offered greater protection for the downstream State than the ALJ's "undue impact" standard suggested. He explained the proper standard as follows:

"'[A] mere theoretical impairment of Oklahoma's water quality standards—i. e., an infinitesimal impairment predicted through modeling but not expected to be actually detectable or measurable—should not by itself block the issuance of the permit. In this case, the permit should be upheld if the record shows by a preponderance of the evidence that the authorized discharges would not cause an actual detectable violation of Oklahoma's water quality standards.' Id., at 117a (emphasis in original)."

On remand, the ALJ made detailed findings of fact and concluded that the city had satisfied the standard set forth by the Chief Judicial Officer. Specifically, the ALJ found that there would be no detectable violation of any of the components of Oklahoma's water quality standards. Id., at 127a–143a. The Chief Judicial Officer sustained the issuance of the permit. Id., at 145a–153a.

Both the petitioners in No. 90–1262 (collectively Arkansas) and the respondents in this litigation sought judicial review. Arkansas argued that the Clean Water Act did not require an Arkansas point source to comply with Oklahoma's water quality standards. Oklahoma challenged the EPA's determination that the Fayetteville discharge would not produce a detectable violation of the Oklahoma standards.

The Court of Appeals did not accept either of these arguments. The court agreed with the EPA that the statute required compliance with Oklahoma's water quality standards, '''[p. 98]''' see 908 F.2d 595, 602–615 (CA10 1990), and did not disagree with the Agency's determination that the discharges from the Fayetteville plant would not produce a detectable violation of those standards. Id., at 631–633. Nevertheless, relying on a theory that neither party had advanced, the Court of Appeals reversed the Agency's issuance of the Fayetteville permit. The court first ruled that the statute requires that "where a proposed source would discharge effluents that would contribute to conditions currently constituting a violation of applicable water quality standards, such [a] proposed source may not be permitted." Id., at 620. Then the court found that the Illinois River in Oklahoma was "already degraded," that the Fayetteville effluent would reach the Illinois River in Oklahoma, and that that effluent could "be expected to contribute to the ongoing deterioration of the scenic [Illinois R]iver" in Oklahoma even though it would not detectably affect the river's water quality. Id., at 621–629.

The importance and the novelty of the Court of Appeals' decision persuaded us to grant certiorari. 499 U.S. 946 (1991). We now reverse.

II
Interstate waters have been a font of controversy since the founding of the Nation. E. g., Gibbons v. Ogden, 9 Wheat. 1 (1824). This Court has frequently resolved disputes between States that are separated by a common river, see, e. g., Ohio v. Kentucky, 444 U.S. 335 (1980), that border the same body of water, see, e. g., New York v. New Jersey, 256 U.S. 296 (1921), or that are fed by the same river basin, see, e. g., New Jersey v. New York, 283 U.S. 336 (1931).

Among these cases are controversies between a State that introduces pollutants to a waterway and a downstream State that objects. See, e. g., Missouri v. Illinois, 200 U.S. 496 (1906). In such cases, this Court has applied principles of common law tempered by a respect for the sovereignty of the States. Compare id., at 521, with Georgia v. Tennessee Copper Co., 206 U.S. 230, 237 (1907). In forging what "may [p. 99] not improperly be called interstate common law," Illinois v. Milwaukee, 406 U.S. 91, 105–106 (1972) (Milwaukee I), however, we remained aware "that new federal laws and new federal regulations may in time pre-empt the field of federal common law of nuisance." Id., at 107.

In Milwaukee v. Illinois, 451 U.S. 304 (1981) (Milwaukee II), we held that the Federal Water Pollution Control Act Amendments of 1972 did just that. In addressing Illinois' claim that Milwaukee's discharges into Lake Michigan constituted a nuisance, we held that the comprehensive regulatory regime created by the 1972 amendments pre-empted Illinois' federal common law remedy. We observed that Congress had addressed many of the problems we had identified in Milwaukee I by providing a downstream State with an opportunity for a hearing before the source State's permitting agency, by requiring the latter to explain its failure to accept any recommendations offered by the downstream State, and by authorizing the EPA, in its discretion, to veto a source State's issuance of any permit if the waters of another State may be affected. Milwaukee II, 451 U.S., at 325–326.

In Milwaukee II, the Court did not address whether the 1972 amendments had supplanted state common law remedies as well as the federal common law remedy. See id., at 310, n. 4. On remand, Illinois argued that §510 of the Clean Water Act, 33 U.S.C. §1370, expressly preserved the State's right to adopt and enforce rules that are more stringent than federal standards. The Court of Appeals accepted Illinois' reading of §510, but held that that section did "no more than [p. 100] to save the right and jurisdiction of a state to regulate activity occurring within the confines of its boundary waters." Illinois v. Milwaukee, 731 F.2d 403, 413 (CA7 1984), cert. denied, 469 U.S. 1196 (1985).

This Court subsequently endorsed that analysis in International Paper Co. v. Ouellette, 479 U.S. 481 (1987), in which Vermont property owners claimed that the pollution discharged into Lake Champlain by a paper company located in New York constituted a nuisance under Vermont law. The Court held the Clean Water Act taken "as a whole, its purposes and its history" pre-empted an action based on the law of the affected State and that the only state law applicable to an interstate discharge is "the law of the State in which the point source is located." Id., at 493, 487. Moreover, in reviewing §402(b) of the Act, the Court pointed out that when a new permit is being issued by the source State's permit-granting agency, the downstream State

"'does not have the authority to block the issuance of the permit if it is dissatisfied with the proposed standards. An affected State's only recourse is to apply to the EPA Administrator, who then has the discretion to disapprove the permit if he concludes that the discharges will have an undue impact on interstate waters. §1342(d)(2).... Thus the Act makes it clear that affected States occupy a subordinate position to source States in the federal regulatory program.' Id., at 490–491."

'''[p. 101]''' Unlike the foregoing cases, this litigation involves not a state-issued permit, but a federally issued permit. To explain the significance of this distinction, we comment further on the statutory scheme before addressing the specific issues raised by the parties.

III
The Clean Water Act anticipates a partnership between the States and the Federal Government, animated by a shared objective: "to restore and maintain the chemical, physical, and biological integrity of the Nation's waters." 33 U.S.C. §1251(a). Toward this end, the Act provides for two sets of water quality measures. "Effluent limitations" are promulgated by the EPA and restrict the quantities, rates, and concentrations of specified substances which are discharged from point sources. See §§1311, 1314. "[W]ater quality standards” are, in general, promulgated by the States and establish the desired condition of a waterway. See §1313. These standards supplement effluent limitations "so that numerous point sources, despite individual compliance with effluent limitations, may be further regulated to prevent water quality from falling below acceptable levels." EPA v. California ex rel. State Water Resources Control Bd., 426 U.S. 200, 205, n. 12 (1976).

The EPA provides States with substantial guidance in the drafting of water quality standards. See generally 40 CFR pt. 131 (1991) (setting forth model water quality standards). Moreover, §303 of the Act requires, inter alia, that state authorities periodically review water quality standards and secure the EPA's approval of any revisions in the standards. If the EPA recommends changes to the standards and the State fails to comply with that recommendation, the Act authorizes the EPA to promulgate water quality standards for the State. 33 U.S.C. §1313(c).

The primary means for enforcing these limitations and standards is the NPDES, enacted in 1972 as a critical part of Congress' "complete rewriting" of federal water pollution '''[p. 102]''' law. Milwaukee II, 451 U.S., at 317. Section 301(a) of the Act, 33 U.S.C. §1311(a), generally prohibits the discharge of any effluent into a navigable body of water unless the point source has obtained an NPDES permit. Section 402 establishes the NPDES permitting regime, and describes two types of permitting systems: state permit programs that must satisfy federal requirements and be approved by the EPA, and a federal program administered by the EPA.

Section 402(b) authorizes each State to establish "its own permit program for discharges into navigable waters within its jurisdiction." 33 U.S.C. §1342(b). Among the requirements the state program must satisfy are the procedural protections for downstream States discussed in Ouellette and Milwaukee II. See §§1342(b)(3), (5). Although these provisions do not authorize the downstream State to veto the issuance of a permit for a new point source in another State, the Administrator retains authority to block the issuance of any state-issued permit that is "outside the guidelines and requirements" of the Act. §1342(d)(2).

'''[p. 103]''' In the absence of an approved state program, the EPA may issue an NPDES permit under §402(a) of the Act. (In these cases, for example, because Arkansas had not been authorized to issue NPDES permits when the Fayetteville plant was completed, the permit was issued by the EPA itself.) The EPA's permit program is subject to the "same terms, conditions, and requirements" as a state permit program. 33 U.S.C. §1342(a)(3). Notwithstanding this general symmetry, the EPA has construed the Act as requiring that EPA-issued NPDES permits also comply with §401(a). That section, which predates §402 and the NPDES, applies to a broad category of federal licenses, and sets forth requirements for "[a]ny applicant for a Federal license or permit to conduct any activity including, but not limited to, the construction or operation of facilities, which may result in any discharge into the navigable waters." 33 U.S.C. §1341(a). Section 401(a)(2) appears to prohibit the issuance of any federal license or permit over the objection of an affected State unless compliance with the affected State's water quality requirements can be ensured. '''[p. 104]'''

IV
The parties have argued three analytically distinct questions concerning the interpretation of the Clean Water Act. First, does the Act require the EPA, in crafting and issuing a permit to a point source in one State, to apply the water quality standards of downstream States? Second, even if the Act does not require as much, does the Agency have the statutory authority to mandate such compliance? Third, does the Act provide, as the Court of Appeals held, that once a body of water fails to meet water quality standards no discharge that yields effluent that reach the degraded waters will be permitted?

In these cases, it is neither necessary nor prudent for us to resolve the first of these questions. In issuing the Fayetteville permit, the EPA assumed it was obligated by both the Act and its own regulations to ensure that the Fayetteville discharge would not violate Oklahoma's standards. See App. to Pet. for Cert. in No. 90–1262, pp. 116a–117a, and n. 14. As we discuss below, this assumption was permissible and reasonable and therefore there is no need for us to address whether the Act requires as much. Moreover, much of the analysis and argument in the briefs of the parties relies on statutory provisions that govern not only federal permits issued pursuant to §§401(a) and 402(a), but also state permits issued under §402(b). It seems unwise to evaluate those arguments in a case such as these, which only involve a federal permit.

'''[p. 105]' Our decision not to determine at this time the scope of the Agency's statutory obligations'' does not affect our resolution of the second question, which concerns the Agency's statutory authority. Even if the Clean Water Act itself does not require the Fayetteville discharge to comply with Oklahoma's water quality standards, the statute clearly does not limit the EPA's authority to mandate such compliance.

Since 1973, EPA regulations have provided that an NPDES permit shall not be issued "[w]hen the imposition of conditions cannot ensure compliance with the applicable water quality requirements of all affected States." 40 CFR §122.4(d) (1991); see also 38 Fed. Reg. 13533 (1973); 40 CFR §122.44(d) (1991). Those regulations—relied upon by the EPA in the issuance of the Fayetteville permit—constitute a reasonable exercise of the Agency's statutory authority.

Congress has vested in the Administrator broad discretion to establish conditions for NPDES permits. Section 402(a)(2) provides that for EPA-issued permits "[t]he Administrator shall prescribe conditions...to assure compliance with the requirements of [§ 402(a)(1)] and such other requirements as he deems appropriate." 33 U.S.C. §1342(a)(2) (emphasis added). Similarly, Congress preserved for the Administrator broad authority to oversee state permit programs:

"'No permit shall issue...if the Administrator...objects in writing to the issuance of such permit as being outside the guidelines and requirements of this chapter.' §1342(d)(2)."

The regulations relied on by the EPA were a perfectly reasonable exercise of the Agency's statutory discretion. The application of state water quality standards in the interstate context is wholly consistent with the Act's broad purpose "to restore and maintain the chemical, physical, and [p. 106] biological integrity of the Nation's waters." 33 U.S.C. §1251(a). Moreover, as noted above, §301(b)(1)(C) expressly identifies the achievement of state water quality standards as one of the Act's central objectives. The Agency's regulations conditioning NPDES permits are a well-tailored means of achieving this goal.

Notwithstanding this apparent reasonableness, Arkansas argues that our description in Ouellette of the role of affected States in the permit process and our characterization of the affected States' position as "subordinate," see 479 U.S., at 490–491, indicates that the EPA's application of the Oklahoma standards was error. We disagree. Our statement in Ouellette concerned only an affected State's input into the permit process; that input is clearly limited by the plain language of §402(b). Limits on an affected State's direct participation in permitting decisions, however, do not in any way constrain the EPA 's authority to require a point source to comply with downstream water quality standards.

Arkansas also argues that regulations requiring compliance with downstream standards are at odds with the legislative history of the Act and with the statutory scheme established by the Act. Although we agree with Arkansas that the Act's legislative history indicates that Congress intended to grant the Administrator discretion in his oversight of the issuance of NPDES permits, we find nothing in that history to indicate that Congress intended to preclude the EPA from establishing a general requirement that such permits be conditioned to ensure compliance with downstream water quality standards.

Similarly, we agree with Arkansas that in the Clean Water Act Congress struck a careful balance among competing policies and interests, but do not find the EPA regulations con- '''[p. 107]''' cerning the application of downstream water quality standards at all incompatible with that balance. Congress, in crafting the Act, protected certain sovereign interests of the States; for example, §510 allows States to adopt more demanding pollution-control standards than those established under the Act. Arkansas emphasizes that §510 preserves such state authority only as it is applied to the waters of the regulating State. Even assuming Arkansas construction of §510 is correct, cf. id., at 493, that section only concerns state authority and does not constrain the EPA 's authority to promulgate reasonable regulations requiring point sources in one State to comply with water quality standards in downstream States.

For these reasons, we find the EPA's requirement that the Fayetteville discharge comply with Oklahoma's water quality standards to be a reasonable exercise of the Agency's substantial statutory discretion. Cf. Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842–845 (1984).

V
The Court of Appeals construed the Clean Water Act to prohibit any discharge of effluent that would reach waters already in violation of existing water quality standards. We find nothing in the Act to support this reading.

'''[p. 108]''' The interpretation of the statute adopted by the court had not been advanced by any party during the Agency or court proceedings. Moreover, the Court of Appeals candidly acknowledged that its theory "has apparently never before been addressed by a federal court." 908 F.2d, at 620, n. 39. The only statutory provision the court cited to support its legal analysis was §402(h), see id., at 633, which merely authorizes the EPA (or a state permit program) to prohibit a publicly owned treatment plant that is violating a condition of its NPDES permit from accepting any additional pollutants for treatment until the ongoing violation has been corrected. See 33 U.S.C. §1342(h).

Although the Act contains several provisions directing compliance with state water quality standards, see, e. g., §1311(b)(1)(C), the parties have pointed to nothing that mandates a complete ban on discharges into a waterway that is in violation of those standards. The statute does, however, contain provisions designed to remedy existing water quality violations and to allocate the burden of reducing undesirable discharges between existing sources and new sources. See, e. g., §1313(d). Thus, rather than establishing the categorical ban announced by the Court of Appeals—which might frustrate the construction of new plants that would improve existing conditions—the Clean Water Act vests in the EPA and the States broad authority to develop long-range, areawide programs to alleviate and eliminate existing pollution. See, e. g., §1288(b)(2).

To the extent that the Court of Appeals relied on its interpretation of the Act to reverse the EPA's permitting decision, that reliance was misplaced. '''[p. 109]'''

VI
The Court of Appeals also concluded that the EPA's issuance of the Fayetteville permit was arbitrary and capricious because the Agency misinterpreted Oklahoma's water quality standards. The primary difference between the court's and the Agency's interpretation of the standards derives from the court's construction of the Act. Contrary to the EPA's interpretation of the Oklahoma standards, the Court of Appeals read those standards as containing the same categorical ban on new discharges that the court had found in the Clean Water Act itself. Although we do not believe the text of the Oklahoma standards supports the court's reading (indeed, we note that Oklahoma itself had not advanced that interpretation in its briefs in the Court of Appeals), we reject it for a more fundamental reason—namely, that the Court of Appeals exceeded the legitimate scope of judicial review of an agency adjudication. To emphasize the importance of this point, we shall first briefly assess the soundness of the EPA's interpretation and application of the Oklahoma '''[p. 110]''' standards and then comment more specifically on the Court of Appeals' approach.

As discussed above, an EPA regulation requires an NPDES permit to comply "with the applicable water quality requirements of all affected States." 40 CFR §122.4(d) (1991). This regulation effectively incorporates into federal law those state-law standards the Agency reasonably determines to be "applicable." In such a situation, then, state water quality standards—promulgated by the States with substantial guidance from the EPA and approved by the Agency—are part of the federal law of water pollution control.

Two features of the body of law governing water pollution support this conclusion. First, as discussed more thoroughly above, we have long recognized that interstate water pollution is controlled by federal law. See supra, at 98–100. Recognizing that the system of federally approved state standards as applied in the interstate context constitutes federal law is wholly consistent with this principle. Second, treating state standards in interstate controversies as federal law accords with the Act’s purpose of authorizing the EPA to create and manage a uniform system of interstate water pollution regulation.

Because we recognize that, at least insofar as they affect the issuance of a permit in another State, the Oklahoma standards have a federal character, the EPA's reasonable, consistently held interpretation of those standards is entitled to substantial deference. Cf. INS v. National Center for Immigrants' Rights, 502 U.S. 183, 189–190 (1991); Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837 (1984). In these cases, the Chief Judicial Officer ruled that the Oklahoma standards—which require that there be "no degradation" of the upper Illinois River—would '''[p. 111]''' only be violated if the discharge effected an "actually detectable or measurable" change in water quality. App. to Pet. for Cert. in No. 90–1262, p. 117a.

This interpretation of the Oklahoma standards is certainly reasonable and consistent with the purposes and principles of the Clean Water Act. As the Chief Judicial Officer noted, "unless there is some method for measuring compliance, there is no way to ensure compliance." Id., at 118a, n. 16 (internal quotation marks omitted; citation omitted). Moreover, this interpretation of the Oklahoma standards makes eminent sense in the interstate context: If every discharge that had some theoretical impact on a downstream State were interpreted as "degrading" the downstream waters, downstream States might wield an effective veto over upstream discharges.

The EPA's application of those standards in these cases was also sound. On remand, the ALJ scrutinized the record and made explicit factual findings regarding four primary measures of water quality under the Oklahoma standards: eutrophication, esthetics, dissolved oxygen, and met- '''[p. 112]''' als. In each case, the ALJ found that the Fayetteville discharge would not lead to a detectable change in water quality. He therefore concluded that the Fayetteville discharge would not violate the Oklahoma water quality standards. Because we agree with the Agency's Chief Judicial Officer that these findings are supported by substantial evidence, we conclude that the Court of Appeals should have affirmed both the EPA's construction of the regulations and the issuance of the Fayetteville permit.

In its review of the EPA's interpretation and application of the Oklahoma standards, the Court of Appeals committed three mutually compounding errors.

First, the court failed to give due regard to the EPA's interpretation of its own regulations, as those regulations incorporate the Oklahoma standards. Instead the court voiced its own interpretation of the governing law and concluded that "where a proposed source would discharge effluents that would contribute to conditions currently constituting a violation of applicable water quality standards, such [a] proposed source may not be permitted." 908 F.2d, at 620. As we have already pointed out, that reading of the law is not supported by the statute or by any EPA regulation. The Court of Appeals sat in review of an agency action and should have afforded the EPA's interpretation of the governing law an appropriate level of deference. See generally Chevron, supra, at 842–844.

Second, the court disregarded well-established standards for reviewing the factual findings of agencies and instead made its own factual findings. The troubling nature of the court's analysis appears on the face of the opinion itself: At least four times, the court concluded that "there was substantial evidence before the ALJ to support" particular findings which the court thought appropriate, but which were '''[p. 113]''' contrary to those actually made by the ALJ. 908 F.2d, at 620, 625, 627, 629. Although we have long recognized the "substantial evidence" standard in administrative law, the court below turned that analysis on its head. A court reviewing an agency’s adjudicative action should accept the agency's factual findings if those findings are supported by substantial evidence on the record as a whole. See generally Universal Camera Corp. v. NLRB, 340 U.S. 474 (1951). The court should not supplant the agency's findings merely by identifying alternative findings that could be supported by substantial evidence.

Third, the court incorrectly concluded that the EPA's decision was arbitrary and capricious. This error is derivative of the court's first two errors. Having substituted its reading of the governing law for the Agency's, and having made its own factual findings, the Court of Appeals concluded that the EPA erred in not considering an important and relevant fact—namely, that the upper Illinois River was (by the court's assessment) already degraded.

As we have often recognized, an agency ruling is "arbitrary and capricious if the agency has...entirely failed to consider an important aspect of the problem." ''Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co.'', 463 U.S. 29, 43 (1983). However, in these cases, the degraded status of the river is only an "important aspect" because of the Court of Appeals' novel and erroneous interpretation of the controlling law. Under the EPA's interpretation of that law, what matters is not the river's current status, but rather whether the proposed discharge will have a "detectable effect" on that status. If the Court of Appeals had been properly respectful of the Agency's permissible reading of the Act and the Oklahoma standards, the court would not have adjudged the Agency's decision arbitrary and capricious for this reason.

In sum, the Court of Appeals made a policy choice that it was not authorized to make. Arguably, as that court sug- '''[p. 114]''' gested, it might be wise to prohibit any discharge into the Illinois River, even if that discharge would have no adverse impact on water quality. But it was surely not arbitrary for the EPA to conclude—given the benefits to the river from the increased flow of relatively clean water and the benefits achieved in Arkansas by allowing the new plant to operate as designed—that allowing the discharge would be even wiser. It is not our role, or that of the Court of Appeals, to decide which policy choice is the better one, for it is clear that Congress has entrusted such decisions to the Environmental Protection Agency.

Accordingly, the judgment of the Court of Appeals is

Reversed.