dissenting:
I respectfully dissent from Part II-B of the opinion. I concur in Part II-A.
We decided this case and filed our opinion in 1993, at 11 F.3d 900. As the majority concedes in footnote 1, the full court voted on whether to go en banc, and decided against it. We were required by Federal Rule of Appellate Procedure 41(a) to issue our mandate seven days after entry of the order denying the petition for rehearing. I do not know why this was not done.
The only justification I can see for us to issue a new decision, contrary to our previous one and contrary to what we said in Oregon Natural Resources Council v. U.S. Forest Service, 834 F.2d 842 (9th Cir.1987), would be a Supreme Court decision to the contrary. When the Supreme Court makes a decision which renders a decision of ours erroneous, we sometimes recall a mandate and revise our disposition in light of the Supreme Court decision. The majority opinion suggests that PUD No. 1 of Jefferson County v. Washington Department of Ecology, — U.S. -, 114 S.Ct. 1900, 128 L.Ed.2d 716 (1994), has changed the law. If it did with respect to an issue dispositive of the case before us, I would not object to recalling our mandate to correct our error. But it does not.
The question before us is whether citizens’ suits may be brought to enforce water quality standards, as opposed to effluent limitations. Jefferson County says nothing about that. Basically, water quality standards say that a body of water should be no more polluted than the standard. Effluent limitations say that a particular discharger may discharge no more than the allowed quantity of pollutants. Jefferson County does not involve a citizens’ suit, says nothing about citizens’ suits, and implies nothing about citizens’ suits. The question the court decided was whether “the state environmental agency, properly conditioned a permit for the project on the maintenance of specific stream flows to protect salmon and steelhead runs.” Id. at -, 114 S.Ct. at 1905. The court decided that EPA’s interpretation, “that oc-tivities, not merely discharges — must comply *991with state water quality standards is a reasonable interpretation of § 401, and is entitled to deference.” Id. at -, 114 S.Ct. at 1909.
If we were to try to tease something out of Jefferson County to assist our decision in this ease, we would do better to focus on the Court’s discussion of the appropriate use of water quality standards, which is to generate limitations for specific activities. Here is the Court’s discussion of the distinction between water quality standards and effluent limitations:
Washington’s Class AA water quality standards are typical in that they contain several open-ended criteria which, like the use designation of the River as a fishery, must be translated into specific limitations for individual projects.
Id. at -, 114 S.Ct. at 1911 (emphasis added); see also id. at -, 114 S.Ct. at 1910 (“States may condition certification upon any limitations necessary to ensure compliance with state water quality standards.” (emphasis added)).
We have said that only permit limitations derived from water quality standards, not water quality standards themselves, are enforceable by citizens’ suits. Oregon Natural Resources Council v. U.S. Forest Service, 834 F.2d 842 (9th Cir.1987).
Thus, effluent limitations may be derived from water quality standards and may be enforced when included in a discharger’s permit. We agree with defendants that it is not the water quality standards themselves that are enforceable in section 1311(b)(1)(C), but it is the “limitations necessary to meet” those standards, or “required to implement” the standards.
Id. at 850 (emphasis added).
WHle our statement in Oregon Natural Resources Council may be dictum as applied to the case at bar, I do not see a good reason for deciding the case at bar inconsistently with what we said in Oregon Natural Resources Council. The Supreme Court has now analyzed the difference between water quality standards and effluent limitations along the same lines in Jefferson County. Judge Ingram’s earlier opinion in the case at bar pointed out that “the authorities generally reject citizen suit standing to enforce water quality standards,” and appellants “have not been able to find a single case in which a court held that citizens’ suits could be used to enforce water quality standards.” Northwest Environmental, 11 F.3d at 907.
The majority reasons that 33 U.S.C. § 1365 allows citizens’ suits for violation of “an effluent standard,” and defines “effluent standard” as “a permit or condition thereof,” and Portland’s permit, in addition to detailed effluent limitations, requires that no wastes be discharged which would violate water quality standards. This reasoning has force, and I am troubled by the difficulty of applying Oregon Natural Resources Council in the face of this logical, literal construction.
Nevertheless, we previously concluded on the basis of analysis of several additional provisions of the statute that it is not the permittee who must comply with the water quality standards, but rather the issuing authority, which has a “duty ... to include in the permit end-of-pipe effluent limitations that will ensure that water quality standards are met.” Northwest Environmental, 11 F.3d at 908. That makes sense in light of what the Supreme Court said in Jefferson County. Congress meant for the issuing authority to decide upon end-of-pipe effluent standards for the permit, which it could derive from water quality standards, when Congress allowed citizens’ suits to enforce permit limitations. It did not mean for citizens’ suits to proceed on the basis of permit violations, where the permittee complied with end-of-pipe discharge limitations but the water still wound up being too polluted. A water quality standard should be deemed to be not among those authorized by the statute for purposes of citizen suit enforcement.
The City of Portland persuasively argues that, if the water quality standard were used as a basis for punishing sewage overflows during rainy weather, then the detailed end-of-pipe discharge limitations in the permit designed for this precise problem would have no practical effect. The City provided evidence that the actual intent of the permit issuing authority and the City of Portland was that the water quality standards lan*992guage was put in for pollutants other than the ones, such as the sewage overflows in rainy weather, known about when the permit was issued. This reading would be consistent with the general principle of treating the specific as overcoming the general. See Karrell v. US, 181 F.2d 981 (9th Cir.1950) (specific provision governs even though general provisions, if standing alone, would include the same subject); International Ass’n of Machinists & Aerospace Workers v. Boeing, 833 F.2d 165, 169 (9th Cir.1987) (same). We should follow the City’s persuasive construction of the permit.
The majority’s argument from legislative history and policy seems to me to have no force. The history shows that because of the ineffectiveness of water quality standards as a pollution limiting device, Congress decided to change the enforcement mechanism to effluent limitations. See Northwest Environmental, 11 F.3d at 909-11.
The majority argues that as a matter of policy, Congress meant to prevent pollution, and citizens’ suits add power to anti-pollution enforcement mechanisms, so there is no reason to deny citizens’ suits enforcement. The first two propositions do not imply the third. There can be too much of a good thing. See, e.g., on other subjects, United States ex rel. Miller v. Greer, 789 F.2d 438, 451-52 (7th Cir.1986) (Easterbrook, J., dissenting); National Labor Relations Bd. v. Walton Manufacturing Co., 289 F.2d 177, 182 (5th Cir.1961) (Wisdom, J., dissenting). There is too much of a good thing when its value is exceeded by the value of other good things available for the same or less cost.
Water quality standards are a useful device for government enforcement authorities (who decided not to prosecute this case against the City of Portland), because they provide standards for effluent limitations and goals toward which enforcement should be aimed. They are too uncertain and amorphous, however, for use against specific polluters. Suppose, hypothetically, that a water quality standard allows for 100 units of a pollutant, upstream and non-point source polluters discharge 50 units, and the downstream discharger is permitted to discharge 50 units. If the upstream and non-point source polluters increase their discharge to 80 units, it does not automatically follow that the downstream discharger should be limited to 20. The burdens of so severe a limitation may exceed the burdens of the extra pollution, or enforcement efforts might more appropriately be directed at the other polluters. In the case at bar, the majority concedes that the social costs of filling the streets and basements of Portland with sewage, or spending between a half billion and $1.2 billion dollars on renovation, are the practical alternatives to tolerating violations of the water quality standards. See Majority Opinion at I-A, II-A. A public authority might rationally decide that filling the streets and basements with sewage is worse than polluting the river with it, and that the citizens of Portland need several years to raise and spend the money necessary to avoid running the sewage into the streets, the basements, and the river.
The reason that this case has not been rendered moot by subsequent changes in the permit which eliminated the basis for the claims is NWEA’s claim for attorneys’ fees and possibly civil penalties. If the private advocacy group which brought this action prevails on the claim that the City should have done something more than it did under its now-obsolete permit, then it will obtain a great deal of money from the citizens of Portland. This incentive is why citizens’ suits may produce too much of a good thing with regard to enforcement.
Suppose, hypothetically, we authorized citizens’ suits to enforce highway speed limits. We can all agree that excessive speeds on the highways are undesirable, and that highway patrol officers are too few to ticket all the drivers who speed. We would get more prosecution, and more court cases, than the prudent exercise of discretion would justify. A police officer might exercise discretion not to ticket drivers less than 10 miles per hour over the limit, because the increased danger was negligible, and the enforcement burden to the officer and the courts great. The officer might, however, ticket a speeder with a makeshift towing rig which looked dangerous. The officer and the judge would be pleased to have only one ticket, not hun*993dreds, to dispose of, and the enforcement expense would be small relative to the reduction in danger. A “public interest advocacy group,” however, would have an incentive to ticket all the automobiles going a few miles per hour over the limit, because the private group, unlike the police officer and the judge, would have a financial incentive to enforce against the large number of minor violators, even though the burdens of enforcement would be very high relative to the improvement in public safety. A zealous concern for safety on the highways would doubtless contribute to doing well by doing good, but there would be too much good done.
In the case at bar, all the precedents indicate that citizens’ suits are not allowable for violations of water quality standards, where those standards are not translated by the permit into effluent limitations. There is no good reason for avoiding application of precedent, and creating new law which allows citizens’ suits for purposes of obtaining attorneys’ fees and penalties for past violations of water quality standards. We should have left our previous decision alone.