This appeal challenges the final power marketing policy of the Southeastern Power Administration (SEPA), a federal power marketing agency within the Department of Energy. Plaintiffs, various municipally-owned electric systems in North Carolina, South Carolina, and Virginia, along with the Cities of Bennettsville and Camden, South Carolina (collectively referred to herein as "ElectriCities"), brought this action in the district court for review of SEPA's final power marketing policy for its Georgia-Alabama system of projects. On cross-motions for summary judgment the district court granted defendants' motion on the alternative grounds that SEPA's actions were nonreviewable because "committed to agency discretion by law," or that SEPA's policy, even if subject to the district court's jurisdiction, was not arbitrary, capricious, or an abuse of discretion. Because we perceive no merit in plaintiffs' appeal, we affirm.

Under the authority of Section 5 of the Flood Control Act of 1944, 16 U.S.C. Sec. 825s, SEPA sells power produced at federal dams to energy consumers.1 The Act requires that SEPA, as delegate of the Secretary of Energy, dispose of the power produced "in such manner as to encourage the most widespread use thereof at the lowest possible rates to consumers consistent with sound business principles." Id. It further requires that preference in the sale of power be given to public bodies and cooperatives, rather than to private entities.

SEPA sells power in a ten-state area, which it has divided into four marketing areas. This appeal involves the final power marketing policy for the Georgia-Alabama system, which includes parts of Georgia, Alabama, South Carolina, North Carolina, Florida, and Mississippi.

SEPA does not own or control any transmission lines. It must depend on others to move the output of its generating facilities to its customers. Also, SEPA supplies only a small part of the needs of its customers. In 1978, its generating capacity amounted to only 21% of the capacity and 10% of the average annual energy of the load of its preference customers in the Georgia-Alabama marketing area. Thus, the demand for power generated by it far exceeds the supply. The power that it sells from the Georgia-Alabama system is "peaking" power, that is, power to meet the peak demands of its customers.

The availability of peaking power from SEPA eliminates the need on the part of its customers to invest in generation facilities to meet the demands for energy made on them during peak periods.

Prior to 1978, SEPA marketed power from its Georgia-Alabama system on an ad hoc basis by negotiating contracts with consumers. In 1978, SEPA adopted a formal procedure for establishing its marketing policies. 43 Fed.Reg. 29186 (1978). Under the new procedure SEPA develops its marketing policies by informal rulemaking with notice and an opportunity to submit comments.

Pursuant to this new procedure SEPA issued a notice of intent to formulate a power marketing policy for the Georgia-Alabama system in 1979. 44 Fed.Reg. 10111 (1979). SEPA received seventy-eight responses to the notice, including written comments from ElectriCities. ElectriCities proposed that once existing contracts for power expired, SEPA should allocate to all interested preference entities proportionately equal allocations.

After the comment period SEPA published a notice of proposed marketing policy for the Georgia-Alabama system. 44 Fed.Reg. 59642 (1979). SEPA subsequently held two public comment fora concerning the proposed policy in January 1980. SEPA received further oral and written comments on its proposed marketing policy, including additional comments from ElectriCities.

On October 1, 1980, SEPA published its notice of the issuance of the Final Power Marketing Policy. The final marketing policy expanded the list of preference customers that would receive some allocation of SEPA power from 150 to 204. Among the added preference customers are several cities that had previously refused SEPA power including several plaintiff cities. Although the final marketing policy expanded the list of preference customers, it also provided that SEPA's pre-existing customers would continue to receive their existing allocations of power after their contracts expired. New customers would only receive power as it became available after the completion of the Richard B. Russell project and the expansion of the Hartwell project both along the Savannah River on the Georgia-South Carolina border. Old and new customers would share the new power proportionately.

In its final marketing policy, SEPA also divided the Georgia-Alabama marketing area into two parts with the Savannah River serving as the boundary between the parts. Customers in the portion of the Georgia-Alabama marketing area that is east of the Savannah River would receive their power from the three projects located on that river. The final marketing policy provided that the eastern portion of the Georgia-Alabama area would consist of the South Carolina Public Service Authority's and South Carolina Electric and Gas Company's service areas and only that part of the Duke Power Company's service area that was within a radius of 150 miles of the Savannah River projects.

Plaintiffs challenged the marketing policy on several grounds. Most important upon appeal are plaintiffs' contentions that SEPA's decisions to treat existing customers preferentially and to exclude potential customers situated more than 150 miles from the Savannah River project are arbitrary and capricious, constitute an abuse of discretion, and are in violation of the Flood Control Act of 1944.

In granting defendants' motion for summary judgment, the district court first ruled that allocation of power among preference customers is a matter "committed to agency discretion by law" and therefore not subject to judicial review. 5 U.S.C. Sec. 701(a)(2). The district court concluded that the Act's command to dispose of power "in such a manner as to encourage the most widespread use thereof" was "simply too vague to supply this Court with a standard by which it can judge the propriety of SEPA's actions." It agreed with the Ninth Circuit that the Act's "most widespread use" language does not provide a standard against which to measure SEPA's marketing decisions, since the language is susceptible to a wide range of divergent interpretations. Santa Clara v. Andrus, 572 F.2d 660, 668 (9 Cir.), cert. denied, 439 U.S. 859, 99 S. Ct. 177, 58 L. Ed. 2d 167 (1978); see also Greenwood Utilities Commission v. Hodel, 764 F.2d 1459, 1464-65 (11 Cir.1985) (holding that section 5 of the Flood Control Act of 1944 provides no law to apply regarding allocation decisions among preference customers). The district court therefore decided that it had no jurisdiction to review SEPA's decision regarding the marketing of power to preference customers.

We agree with the district court. Agency actions are presumptively subject to judicial review except where Congress manifests its intent to preclude such review. Abbott Laboratories v. Gardner, 387 U.S. 136, 140-41, 87 S. Ct. 1507, 1510-11, 18 L. Ed. 2d 681 (1967); but see Heckler v. Chaney, --- U.S. ----, 105 S. Ct. 1649, 84 L. Ed. 2d 714 (1985) (agency nonenforcement decisions presumptively nonreviewable). However, the Administrative Procedure Act (APA) provides that agency action is nonreviewable if it is "committed to agency discretion by law." 5 U.S.C. Sec. 701(a)(2). The exception is narrow. It applies only "in those rare instances 'where statutes are drawn in such broad terms that in a given case there is no law to apply.' " Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 410, 91 S. Ct. 814, 820, 28 L. Ed. 2d 136 (1971) (quoting legislative history of the APA, S.Rep. No. 752, 79th Cong., 1st Sess., 26 (1945)). Despite harsh criticism of the no-law-to-apply doctrine, the Supreme Court has reaffirmed the doctrine in Heckler v. Chaney, --- U.S. ----, 105 S. Ct. 1649, 84 L. Ed. 2d 714 (1985).2

We agree with the district court that the phrase "most widespread use" does not limit SEPA's discretion to act in the manner which is challenged by plaintiffs and thus, in this case, there is no law to apply. Like the district court and the Ninth and Eleventh Circuits we conclude that taken on its own terms the statutory language is too vague to provide a standard by which a court can review SEPA's decisions. As the Ninth Circuit has noted, the phrase "most widespread use" might mean most geographically widespread distribution of power, distribution to the most diversified mix of ultimate consumers, or distribution to preference customers that reach the greatest number of ultimate consumers. Santa Clara, 572 F.2d at 668. Nor do the scant references in the legislative history and in contemporaneous administrative materials to the Act's provision for power marketing help plaintiffs. At most we are able to glean from these materials a congressional purpose to avert monopolization of power3 and to foster the general goals of rural electrification.4 Neither purpose supplies law to apply in this particular case.

We are, of course, mindful of the Supreme Court's warning that instances of agency action that fall within the Section 701(a)(2) exception of the APA thereby escaping judicial review are rare. Nonetheless, we think the conclusion inescapable that SEPA's marketing decisions among preference customers under Section 5 of the Flood Control Act of 1944 are "committed to agency discretion by law." We therefore conclude that the district court properly held that it had no jurisdiction to review SEPA's allocation decisions.

It is, of course, true that "agency action [which] is committed to agency discretion by law" is not completely shielded from judicial review. Most recently, Justice Brennan stated in his concurring opinion in Chaney that courts may review agency actions for certain types of errors that fall within the APA's Section 701(a)(2) exception to judicial review. For example, an agency decision that violates a statutory or constitutional command or is prompted by a bribe is not immune from judicial review even when a lawful exercise of an agency's discretion has that immunity. Chaney, --- U.S. at ----, 105 S. Ct. at 1659-60 (Brennan, J., concurring). In Garcia v. Neagle, 660 F.2d 983, 988 (4 Cir.1981), cert. denied, 454 U.S. 1153, 102 S. Ct. 1023, 71 L. Ed. 2d 309 (1982), we similarly said that "even where action is committed to absolute agency discretion by law, courts have assumed the power to review allegations that an agency exceeded its legal authority, acted unconstitutionally, or failed to follow its own regulations," but they "may not review agency action where the challenge is only to the decision itself." See also Ness Investment Corp. v. United States Department of Agriculture, 512 F.2d 706, 714 (9 Cir.1975); Scanwell Laboratories, Inc. v. Shaffer, 424 F.2d 859, 874 (D.C.Cir.1970); 5 K. Davis, Administrative Law Treatise Sec. 28.2 (2d ed. 1984).

On this basis, plaintiffs also challenge the marketing policy, contending that in developing the policy SEPA improperly considered existing customers' prior political support for SEPA and for funding of Federal power projects. Even where an agency's action is committed to the agency's discretion, a court should hear a claim that an agency rested an action on "considerations that Congress could not have intended to make relevant" such as the political considerations alleged by plaintiffs. Littell v. Morton, 445 F.2d 1207 (4 Cir.1971) (quoting Wong Wing Hang v. INS, 360 F.2d 715, 719 (2 Cir.1966)).5 While the district court failed to address this part of plaintiffs' claim, our review of the record persuades us that it is lacking in merit.

To show improper political influence, plaintiffs rely on two documents contained in the record and the testimony of the Administrator of SEPA given at the Public Comment Forum held on January 10, 1980. We consider this evidence seriatim.

The first reference is to language contained in the document entitled "Power Marketing Policy Considerations" prepared by SEPA and dated October 1977. In pertinent part, the document discussed treatment of existing customers and proceeded from the premise that if all potential customers were on an equal footing the allocation of existing power would normally be in relation to the magnitude of their power loads. However, the report continued, not all potential customers are on equal footing; some were then purchasing power and desired to purchase more; some initially declined to purchase power but desire to purchase now; some are outside of SEPA's selected marketing area and have not been offered the opportunity to purchase power. Since existing customers now purchase all of the available energy, any allocation to any of the other two groups would take power away from existing customers. Then, with regard to existing customers, the report stated:

In general these customers are well organized and have supported the Federal power program over a long period of time. In many cases they have actually championed the projects and were instrumental in obtaining congressional authorization and funding of the projects. It is believed that these customers and their political representatives will be extremely vocal and will resist to the utmost the adoption of any power allocation procedure which provides that some power which they now purchase will be taken away and sold to others.

Immediately following this statement, the report said, with respect to the other two classes of potential customers:

On the other hand potential customers in groups 2 and 3 are adamant that they will receive an allocation of power at the earliest possible moment. One possible consideration is the development of an expanded power product made possible by purchasing energy from the private utilities in sufficient quantities so that this energy together with the project capacity now sold to the private utilities would provide an additional Federal power supply available for sale to preference customers.

The second document, dated April 30, 1980, is the written staff response to statements made by customers and potential customers at the public fora and the pertinent portion is the staff comment to the suggestion that all preference entities should receive an allocation of power and "there should be no different treatment accorded existing preference customers over preference entities not now served [with certain exceptions not pertinent here]." The staff response was that SEPA's resources, even projected to the completion of new facilities in 1985, could meet only 6% of the capacity requirements and less than 5% of the energy requirements of preference entities in the merged marketing area; that "[s]ince its creation in 1950, SEPA has progressively enlarged its marketing areas and increased the number of preference customers served [but] [o]nce a supplier-customer relationship was established, SEPA has never unilaterally terminated such a relationship [nor] has SEPA ever reduced an allocation once established." The report then stated:

The advocates of ultimate equal treatment as far as existing power is concerned and exclusive interim rights to new power sources knowingly take a very provocative position. These public bodies are not only located remote from the power sources, old or new, but have shown little or no support or interest in development of these power sources while many of the existing customers have been longtime proponents of the projects giving of themselves and their treasure in bringing the resource development to fruition. It would appear that an appropriate course of action for these remotely located public entities to pursue, if they desire Federal projects to be a source of power supply for them, would be to become active in the location, promotion, authorization and construction of such projects. The National Hydropower Study is presently underway for the express purpose of identifying potential new projects. It is the price that others have paid.

Staff Evaluation Committee, Staff Evaluation of Proposed Power Marketing Policy, Georgia-Alabama System of Projects at 6 (April 30, 1980).

The testimony was that of Harry F. Wright, Administrator of SEPA, given at the Public Comment Forum held on January 10, 1980. There, in the course of questioning about the proposed policy, the following was elicited:

MR. HORWOOD: I guess my question is, was there past support for the Federal Power Marketing Program and SEPA a factor that was considered or was it a factor that was basically ignored?

MR. WRIGHT: Well, I think we tried to consider all of the factors. I mean, you can't just say everything is ignored, and it's hard to say that because of this factor we made this decision. But, of course, that is a claim, and, I guess the fact that some of the preference customers have, in the past, substantially supported the projects and, as a result of that, they think they do have a right to the power, that is one factor that's considered as well as others.

MR. BELL [General Counsel of SEPA and Forum Chairman]: I'll add this, Mr. Horwood, though. We have offered power to customers who have offered no support to us at all, including customers in North Carolina.

We cannot read this evidence as supporting a charge of improper political influence. At most, it is a combination of two factors. First, it is a recognition that if the proposal advanced by some potential customers that allocations of power to existing customers be reduced, there would be vigorous opposition and recriminations from existing customers. Second, it is recognition of the validity of a marketing approach employed widely in private business, namely, that an existing customer, as a result of its past custom, is entitled to protection if not preferential treatment with regard to future business. Indeed, the latter had been the policy of SEPA for over thirty years, and we see nothing wrong in the continuation of that policy, least of all that it constitutes evidence of a tainting, improper political influence. Thus, we hold that the evidence of improper political influence is insufficient as a matter of law to warrant consideration of it by the district court.

Because the district court properly held that SEPA's allocation decisions among preference customers was committed to agency discretion by law, and therefore not subject to judicial review, it was unnecessary for the district court to measure SEPA's marketing policy against the standard of review set forth in 5 U.S.C. Sec. 706(2)(a).6 We therefore need not discuss the district court's alternative holding in detail. We are persuaded by the district court's reasoning, however, and agree with it that even under this standard of review, SEPA's policy is neither unlawful nor invalid. Because the "most widespread use" language is subject to divergent meanings we cannot say that SEPA's policy fails to accord with the Flood Control Act of 1944. Nor did SEPA abuse its discretion or act in an arbitrary or capricious manner. Review under this standard does not permit a court to substitute its judgment for an agency's. The district court properly ruled that SEPA had stated a rational basis for its decision to retain the present allocations of pre-existing customers and to impose the 150 mile radius limitation. The former was supported by SEPA's concern about the "radical impact upon existing customers and SEPA's marketing program" that ElectriCities' alternative allocation proposal would have, and the latter by SEPA's explanation about its concerns regarding its limited resources.

Affirmed.