Petition to review an order of the Governors of the United States Postal Service (Governors) which ordered new postal rates and fees to take effect on February 17, 1985 pursuant to the Postal Reorganization Act of 1970, 39 U.S.C. Secs. 101 et seq. (1976) (the Act). Petitioners and intervenors present numerous challenges to the Governors' decision. These challenges include whether certain of the rates implemented are supported by substantial record evidence. The structure of the Postal Service and the procedure by which the Governors approve postal rates have been examined extensively elsewhere, see National Ass'n of Greeting Card Publishers v. United States Postal Service, 462 U.S. 810, 103 S. Ct. 2717, 77 L. Ed. 2d 195 (1983); Time, Inc. v. United States Postal Service, 685 F.2d 760 (2d Cir.1982) [hereinafter cited as Time I ]; Newsweek, Inc. v. United States Postal Service, 663 F.2d 1186 (2d Cir.1981), aff'd, National Ass'n of Greeting Card Publishers, 462 U.S. 810, 103 S. Ct. 2717, 77 L. Ed. 2d 195 and we will provide only a brief review.

The United States Postal Service was established by the Act as an independent agency under the management of an eleven member Board of Governors (Governors). The Act also established a five member Postal Rate Commission (PRC) which is independent of the Board. See 39 U.S.C. Secs. 201, 202, 3601. When the Board deems it to be in the public interest, it may request the PRC to recommend a new postal rate schedule. Id. Sec. 3622(a). Upon receiving such a request, the PRC holds hearings, id. Sec. 3624(a), and formulates a schedule of rates, id. Sec. 3624(d), in accordance with nine factors listed in Sec. 3622(b). Upon receipt of the PRC's recommended rates, the Governors have several options: they may approve, allow under protest, reject or modify the rates. Id. Sec. 3625(a). Once the Governors order new rates into effect, that decision may be appealed to any Court of Appeals which may affirm the decision or order the entire matter returned for further consideration, but may not modify the decision. Id. Sec. 3628.

On November 10, 1983, the Postal Service filed a request for changes in postal rates with the PRC. The PRC then established Docket No. R84-1, the sixth major postal rate proceeding since the Act was enacted in 1970, and held hearings in which some 71 parties intervened. The record of these hearings comprises over 35,000 pages and includes the testimony of over one hundred witnesses. The PRC issued its Opinion and Recommended Decision on September 7, 1984 in accordance with a ten month statutory deadline. The Governors then approved the recommended rates and fees as "consistent with the policies of the Postal Reorganization Act and supported by adequate record evidence." Joint Appendix at 2. The new rates were placed into effect on February 17, 1985.

Petitions for review were filed by five parties. Nine parties intervened in this court in support of the Postal Service's approval of the recommended rates. We now turn to the arguments presented by petitioners in support of their respective contentions.

The American Newspaper Publishers Association (ANPA) contends that the decision of the Governors cannot be reviewed by this court because the Governors did not make detailed findings and failed to provide a sufficient explanation of the basis for adopting the PRC recommendation. Furthermore, ANPA claims that the Governors did not expressly adopt the PRC recommendation and in fact expressed reservations about aspects of the PRC's conclusions.

The Act divides the responsibility for ratemaking between the Governors and the PRC and specifies the procedures by which the Governors approve postal rates. Time, Inc. v. United States Postal Service, 710 F.2d 34, 36 (2d Cir.1983) [hereinafter cited as Time II ]; see 39 U.S.C. Secs. 201-02, 3601. "While the [Governors are] is entrusted with the power 'to establish ... reasonable and equitable rates of postage and fees for postal services,' see 39 U.S.C. Sec. 3621, its actual rate-making authority, with a few exceptions, is limited to approving, rejecting, modifying, or allowing under protest recommendations of the PRC, see 39 U.S.C. Sec. 3625(a)." Newsweek, 663 F.2d at 1190. The Governors decision must include an estimate of anticipated revenue and a statement of explanation and justification, 39 U.S.C. Sec. 3625(e). In addition, postal rates must provide sufficient revenue so that total revenues equal as nearly as practicable total costs of the Postal Service. Id. Sec. 3621; see Time II, 710 F.2d at 36-37. The Governors decision may be set aside only if it is found by a reviewing court to be arbitrary and capricious, an abuse of discretion, not in accordance with law, or unsupported by substantial evidence. 39 U.S.C. Sec. 3628; 5 U.S.C. Sec. 706.

We hold that the Governors decision meets the statutory criteria for approval of the PRC's recommendations. We note that the scope of our review has been limited to determining whether a " 'reasonable mind' [would] find the evidence on which the agency relied to be 'adequate' to support its conclusion." Newsweek, 663 F.2d at 1210. As long as the path by which the agency reached its decision is discernible, see Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 286, 95 S. Ct. 438, 442, 42 L. Ed. 2d 447 (1974), its decision should be upheld.

In their decision, the Governors stated that:

Pursuant to our authority under 39 U.S.C. Sec. 3625 and recognizing the Commission's expertise and statutory responsibility, we approve the recommended rates and fees because we believe that they are consistent with the policies of the Postal Reorganization Act and supported by adequate record evidence. We believe that they will enable the Postal Service to break even during the test year, in accordance with the mandate of the Act.

Governors' Decision at 2. The Governors' decision also estimated anticipated revenues and discussed the rates for each of the classes of mail pursuant to Sec. 3625(e). Id. at 2-15. The PRC evidentiary record on which the Governors relied was voluminous; the PRC opinion itself was 609 pages and was supplemented by lengthy appendices. Our decisions in Time I and Time II are inapposite to the issues raised before us. In those decisions, the Governors sought to modify the PRC recommendation. We held that the Governors must provide this court with a justification of the modifications for each class of mail so that we could determine whether the Governors' modifications were supported by substantial evidence. See Time I, 685 F.2d at 772. Here, the Governors are not modifying the PRC's recommendation.

There is no merit to the contention that the September 7, 1984 PRC decision was not adopted because the Governors expressed reservations about some aspects of that decision. In National Ass'n of Greeting Card Publishers v. United States Postal Service, 607 F.2d 392 (D.C.Cir.1979), cert. denied, 444 U.S. 1025, 100 S. Ct. 688, 62 L. Ed. 2d 659 (1980), the court held that "[t]he mere fact that the [Governors] found the PRC decision to be not 'totally acceptable' or 'less than wholly persuasive' does not vitiate the [Governors'] approval.... We decline to add the judicial veneer of a non-statutory requirement that the [Governors] find a PRC decision 'wholly acceptable' in order to approve it." Id. at 420-21. In our view, the Governors' approval and adoption of the PRC decision meet the statutory requirements of the Act.

We begin our analysis of the PRC's recommended rates with a discussion of the rates implemented for third-class bulk rate regular mail (hereinafter BRR). We emphasize that in tracing the path followed by the PRC in recommending, and the Governors in adopting, the rates in question, we may not substitute our judgment for that of the agencies designated by law to carry out the ratemaking function. We reaffirm that "the rate-setting power is committed to agency discretion, subject to broad guidelines rather than specific, rigid formulae." Newsweek, 663 F.2d at 1198 (emphasis in original).

Before beginning our detailed analysis, it is necessary to provide an explanation of some of the terms and concepts used by the PRC in this ratemaking docket. The final BRR rates, as with the other rates set in this docket, were determined primarily upon the basis of "cost coverage." That is, in compliance with the statutory requirement embodied in 39 U.S.C. Sec. 3622(b)(3), each class or type of mail must "bear the direct and indirect postal costs attributable to that class or type plus that portion of all other costs of the Postal Service reasonably assignable to such class or type." This means "[f]irst, [that] all costs that in the judgment of the Rate Commission are the consequence of providing a particular class of service must be borne by that class." National Ass'n of Greeting Card Publishers, 462 U.S. at 833, 103 S.Ct. at 2732. These costs are hereinafter termed "attributable costs." Next, "[t]he Rate Commission is to assign remaining costs reasonably on the basis of the other eight factors set forth by Sec. 3622(b)." Id. at 834. These costs are termed "institutional costs," since they are not the direct consequence of providing any one type of mail service. Cost coverage, then "reflects the contribution to institutional costs provided by a class as a percentage [of] the amount of attributable costs incurred by that class. Thus a class with [attributable] costs of 10, which pays rates designed to cover [attributable] costs and provide a contribution [to institutional costs] of 5, is said to have a cost coverage of 150 per cent." PRC Opinion at 261. Postal rates reflect both attributable costs and the contribution to institutional costs.

Three parties join in the attack on the BRR rates set by the PRC. Petitioner Direct Marketing Association (DMA) is an organization of mailers and takes the position that the cost coverage assigned to BRR mail is too high and that therefore the BRR rates are too high. Petitioner ANPA is an association representing more than 90 percent of the daily newspapers published throughout the United States. They are heavy users of the Postal Service, but also compete with the Postal Service for the business of advertisers who may utilize either newspapers or BRR mail for the saturation distribution of advertising circulars. The thrust of ANPA's arguments regarding BRR is that the BRR rates are too low; higher rates would discourage advertisers from using the mail and, presumably, would encourage them to use the distribution services of ANPA's members instead. Finally, the Coalition of Non-Postal Media (CNPM) is an association of private delivery companies and weekly newspapers that compete with the Postal Service and with ANPA's members for the carriage of hard-copy advertising and which, like ANPA's members, believe themselves competitively disadvantaged by BRR rates that are too low. In sum, DMA seeks to lower the BRR rates, while ANPA and CNPM, being competitors of the Postal Service (as well as of each other), seek to raise postal rates in an effort to gain some of the Postal Service's business.

All three parties described above attack the BRR rates on the ground that the PRC's determination of the appropriate cost coverage for BRR mail was inappropriate. Each of them argues that the PRC's sole explanation for the BRR cost coverage was contained in the following paragraph:

As to cost coverage, we recommend 146 percent, which is lower than our recommended First-Class cost coverage of 159 percent. We thus return to the relative relationship that existed between First- and Third-Class before Docket No. R80-1, when our use of the Service-Related Costs concept (SRC) was an acknowledged influence on our departure from the traditional relationship.

PRC Opinion at 464 (quoted in DMA Brief at 29; ANPA Brief at 26; CNPM Brief at 47). ANPA and CNPM also claim that even if such a one-paragraph explanation could be sufficient, this one was not, because there exist no means by which the "traditional relationship" adverted to can be ascertained.

Regarding the "traditional relationship" referred to by the PRC, we consider it clear that what was meant was a cost coverage for BRR that was lower than that for First-Class and the systemwide average; the PRC had, as it noted, departed from this "traditional" relationship in Docket No. R80-1, when it de-emphasized cost coverage as a means of setting appropriate rates in favor of "service-related costs." See PRC Opinion in Docket No. R80-1 at 223, 226-27, 449. After the decision of the Supreme Court in National Ass'n of Greeting Card Publishers, 462 U.S. 810, 103 S. Ct. 2717, 77 L. Ed. 2d 195, affirming this Court's decision in Newsweek, 663 F.2d 1186, it is clear that the PRC is not bound to apply "service-related costs" principles and was accordingly free to return to the "traditional" relationship between cost coverage for BRR and the systemwide average cost coverage.

In our view, the PRC offered an adequate explanation for the BRR cost coverage, and hence for the resulting rates, that adequately addressed each of the necessary statutory factors. Although the PRC explanation is at times less than a model of clarity, we emphasize that all we "may properly do is to consider whether the [PRC] did take into account all the relevant factors and no others." Association of Am. Publishers, Inc. v. Governors of the United States Postal Service, 485 F.2d 768, 774-75 (D.C.Cir.1973) (quoted in Newsweek, 663 F.2d at 1200). We reiterate that courts should "uphold a decision of less than ideal clarity if the agency's path may reasonably be discerned." Bowman, 419 U.S. at 286, 95 S.Ct. at 442. We believe the PRC's path to be reasonably discernible and therefore within the ambit of Bowman. Accordingly, we turn to a discussion of the relevant factors which the PRC was bound to consider, and describe the path taken by the agency in applying them.

Title 39, U.S.Code, section 3622(b), sets out nine factors, in addition to "the policies of this title," that must be considered in setting postal rates. In this docket, the procedure followed by the PRC was to assign BRR a cost coverage of 146 percent, and then to examine the testimony and evidence presented, to determine whether any change in the assigned cost coverage was warranted. DMA criticizes this procedure as pulling a cost coverage figure "out of thin air"; we do not, however, perceive any statutory ban on using a judgmental approach to assigning institutional costs. Indeed, as the Postal Service points out, ratemaking is an iterative process. When appropriate relative rate levels have been established, rates must be designed, and their revenue and cost implications must be considered. Because the rates affect mail volume, and total revenue must equal total cost, each change in proposed rates requires a recalculation of volume, revenue and cost. Whether this approach be judgmental, or DMA's econometric concept, the dynamic nature of the ratemaking process would make it difficult, at the very least, to explain why a given cost coverage is 146 percent and not, for example, 148 percent.

In any event, the PRC's task in assigning institutional costs is to assign them "reasonably on the basis of the ... eight factors [in addition to subsection (b)(3) ] set forth by Sec. 3622(b)." National Ass'n of Greeting Card Publishers, 462 U.S. at 834, 103 S.Ct. at 2732. Our task is only to ascertain that all of these factors were taken into account. Thus, if the PRC justified its cost coverage on the record, with reference to the factors contained in Sec. 3622, this court may not disturb the rates adopted by the agency. It is our view that the PRC did so, and we turn now to an examination of the means by which it did.

The first consideration embodied in the statute is that the rates must be "fair and equitable." As noted, the PRC first assigned BRR a cost coverage of 146 percent, which fell between the cost coverage assigned to First-Class mail, 159 percent, and that assigned to parcel post, 116 percent. Then, in order to determine whether this cost coverage and the rates that would result from it were "fair and equitable" so as to comply with subsection (b)(1), the PRC applied the remaining factors of the Act, and also considered the testimony of the witnesses. Some of the factors that the Act requires the PRC to consider in setting recommended rates--such as this one--are not susceptible of exact quantification. Still, postal ratemaking is a function delegated to the expertise of the PRC, and we will not find an approach by it in performing its function to be unacceptable, as long as the PRC adequately justifies its decision on the basis of the Act's requirements.

This element of the Act requires the PRC to consider, in assigning institutional costs, "the value of the mail service actually provided ... to both the sender and the recipient, including but not limited to the collection, mode of transportation, and priority of delivery." The participants before the PRC focused on the quality of the service provided; thus, for example, ANPA's witness Chown emphasized improvements in the quality of delivery that had allegedly taken place since the last rate case in 1980, while the testimony of witnesses sponsored by the mailers, e.g. DMA, emphasized lengthy delivery periods, inconsistent delivery, nondelivery, and general failure of the Postal Service to meet published delivery standards. See PRC Opinion at 516. The PRC concluded that the arguments of ANPA witness Chown did not warrant an increased cost coverage for BRR, in light of the countervailing testimony of the mailers. Id.

The PRC also examined whether a lower cost coverage for BRR was warranted. It concluded that a decreased cost coverage was not warranted on the present record, adopting Postal Service witness Allen's view that the value of service for the BRR subclass was "somewhat mixed," Joint Appendix at 270-71. This view is based upon the observation that despite the seemingly low quality of service in BRR mail, and the resulting high price elasticity of demand, BRR remains attractive to mailers, as evidenced by the rapid increase in volume of BRR mail. Id. That the PRC shared this view is evidenced by its statement that "[w]itness Allen correctly pointed out that [BRR] has features which make it attractive to mailers, despite the low value associated with its speed of delivery." PRC Opinion at 297-98. Accordingly, the PRC did not deviate from its recommended cost coverage of 146 percent on the basis of value of service considerations.

DMA devotes much of its argument to an attack on the PRC's assessment of subsection (b)(2). The thrust of its argument is that the PRC is required by the Act to use "value-of-service pricing," one variant of which is known as the "inverse elasticity rule," in assigning institutional costs. The argument is based on the Report of the President's Commission on Postal Organization (Kappel Commission Report), the relevant portions of which relied heavily upon a report prepared by Foster Associates (Foster Associates Study). The Foster Associates Study, according to DMA, recommended that value of service pricing play a central role in postal ratemaking, and, again according to DMA, Congress adopted this approach in the Act. Value of service pricing involves assigning institutional costs "to the classes [of mail] in accordance with mark-ups inversely proportional to their elasticities of demand." DMA Brief at 19. This, it is asserted, will promote economic efficiency.

We disagree with the argument that Congress intended relative demand to be the benchmark for the assignment of institutional costs. Rather, it is clear that no single factor was intended by Congress to be the "primary" factor in making the assignments. Newsweek, 663 F.2d at 1200. All of the factors must be considered, id., and as we shall show, they were. Indeed, "that the Senate Report refers to all of the specific factors of section 3622(b) as 'requirements' is a strong indication that no one factor was intended to carry more weight than any other." Id. at 1199. Even within subsection (b)(2), we cannot agree that Congress intended value of service pricing, as defined by DMA, to be the paramount consideration. Rather, value of service as embraced by subsection (b)(2) "embodies two concepts: what the customer gets and what he will pay." Kappel Commission Report at 132, reprinted in DMA Brief app. A at 12a. DMA's approach focuses on what the customer will pay, and largely ignores what the customer gets. The latter consideration, referred to by the Postal Service as "intrinsic value," involves the criteria actually set forth in subsection (b)(2): collection, mode of transportation, and priority of delivery. These, as discussed above, the PRC did consider. Moreover, the PRC did apply quantitative factors, although not in the manner DMA wished it to: as the PRC noted, the service-related cost (SRC) method

led us to believe that many quality factors had been attributed or assigned to the classes and subclasses because of their cost effects and consequently, that value of service should receive less emphasis. Since we are not using SRC in this case, the same reasoning does not apply.... [T]he Commission's current method for applying all factors judgmentally, reaches the result [that the value-of-service advocates] acknowledg[e] is both predictable and permissible; that is, prices that are not strictly in accord with Ramsey principles.

PRC Opinion at 292-93, 313.

Moreover, we agree with the Postal Service that DMA's approach places unwarranted emphasis on relative demand for postal services. While there may be merit to the use of a relatively more quantitative approach to the assignment of institutional costs than was applied by the PRC in this docket, this judgment is committed to the discretion of the PRC. As the PRC noted, "[w]hile the Commission may be free to use quantitative methods, it is not required to do so." Id. at 297. The judgment as to what combination of methods to use, and the weight to be accorded each, is a function of agency discretion that we are not inclined to disturb herein.

Subsection (b)(3) requires "that each class of mail or type of mail bear the direct and indirect postal costs attributable to that class or type." The PRC concluded with respect to this requirement that the 146 percent cost coverage would enable BRR mail to recover all its attributable costs. PRC Opinion at 524. We see no reason to disturb the PRC's conclusion.

The PRC considered these two statutory factors together, an approach that makes sense because subsection (b)(4) requires the PRC to consider the effect of rate increases upon the public, business mailers, and competitors of the Postal Service, while subsection (b)(5) requires consideration of available alternatives to the Postal Service for sending non-letter mail.

Without specifically mentioning impact on mailers, the PRC determined that the 146 percent cost coverage was appropriate, because although the 146 percent figure represents a lower than average cost coverage (the systemwide average is 152 percent), the rate increase for BRR, 13 percent, was higher than the average (the average rate increase was 9 percent). See PRC Opinion at 1, 4, 315. Although impact on mailers was not specifically mentioned, it seems clear to us that this is what the PRC had in mind. It is mailers, including the general public and businesses, who directly benefit from a mitigation of rate increases; the PRC's statement that cost coverage would not be raised further because the rate increase was already higher than the average clearly was made in view of the impact of an even higher rate increase upon mailers.

ANPA, a competitor, assails the PRC's recommended rate on the additional ground that the PRC inadequately considered unit contribution as a measure of a reasonable rate. The PRC, however, made it clear in its Opinion that it would primarily rely upon cost coverage as the measure of appropriate rates, using unit contribution only as an adjunct to review cost coverage relationships. Id. at 261. The unit contribution for BRR mail was calculated, see id. Appendix G, Schedule 1, at 1; thus it seems clear that the PRC was aware of the unit contribution for BRR mail, although unit contribution was not expressly mentioned elsewhere. That it did not expressly advert to unit contribution is understandable, given its limited role as a measure of an appropriate postal rate; moreover, given the wide range of unit contributions between different classes of mail, ranging from 2.9cents for Second-Class regular rate to almost $8.00 for Express Mail, id., the supplementary role of unit contribution does not seem unreasonable. ANPA's proposal to the PRC, that the BRR contribution equal 60 percent of that for First-Class mail, which was the relationship that obtained in Docket No. R80-1, would have resulted in a rate increase of 48.7 percent for BRR. Joint Appendix at 606. The PRC declined to adopt this proposal, and in doing so, we conclude that the PRC properly exercised its discretion in evaluating unit contribution and ANPA's proposal.

CNPM devotes most of its argument to an attack on the PRC's consideration of issues relating to competition. Its position is that the PRC set BRR cost coverage too low, because it inadequately addressed the impact of BRR rates upon private-sector competitors of the Postal Service.

At the hearings before the PRC, CNPM's witnesses and discovery responses solicited from its members tended to show that they had suffered financial harm because of low BRR rates that were put into effect in 1981. CNPM sought relief from what it termed "anticompetitive postal rates." PRC Opinion at 519.

In the present appeal, CNPM claims that the PRC failed "to give meaningful consideration to the effect of the approved BRR rates upon competitors engaged in the private, non-postal delivery of advertising material." CNPM Brief at 11. This alleged failure is asserted to violate subsection (b)(4). We do not agree.

First, it is clear that the PRC considered CNPM's evidence, but rejected CNPM's conclusions. See PRC Opinion at 519. In addressing the effect of rates upon competitors, the PRC engaged in a three-phase analysis. First, the PRC examined the question whether the BRR rates were, of themselves, anticompetitatively low. Id. at 521-22. In making this determination, the PRC examined the relationship of price to attributable costs, a variation of the Areeda-Turner predatory pricing test. Id. Because the BRR rates would recover attributable costs, pursuant to subsection (b)(3), the PRC concluded that no presumption of predatory pricing applied to the BRR rates. CNPM asserts that the PRC erred in considering the Areeda-Turner test, since this case is not an antitrust matter. It is, however, clear that ratemaking agencies may consider antitrust principles in assessing impact upon competition. See Northern Natural Gas Co. v. FPC, 399 F.2d 953, 959 (D.C.Cir.1968). Even if, as CNPM asserts, the test was not applied with the greatest possible degree of precision, we note that there is no requirement that the test be applied at all. We conclude that the PRC did not err by examining whether the BRR rates were, on their face, anticompetitive.

Next, the PRC examined the evidence proffered by CNPM of harm to its members. It concluded that while some firms were apparently able to compete, others were not. PRC Opinion at 522. Some firms were able to compete on a piece-by-piece basis, but not with so-called "shared mail," a product marketed by entities which utilize BRR mail to send assembled sets of advertising pieces to customers. Thus, those members were successful in competing in terms of prices, but unsuccessful in competing with a shared mail product. These members, however, clearly have the ability to assemble advertising sets, and as CNPM itself noted, to do so would seem to be a practical necessity. These firms, then, are at least theoretically able to compete in this new product submarket, because the ability to create the new product is independent of postal rates. Id. The PRC thus concluded that the BRR cost coverage need not be adjusted upward to account for the impact on private delivery companies. Id.

The PRC went on, however, to consider whether any weight increments within the BRR subclass presented a particular competition problem, because CNPM had indicated a particular interest in pricing at the three to four ounce level and above. The PRC found no basis in the record to conclude that the BRR rates were improperly skewed with respect to those weight levels. Id. at 523.

In addition to examining CNPM's arguments, as outlined above, the PRC also examined the impact upon other competitors. For example, ANPA's members, while also claiming that BRR rates were too low, emphasized competitive issues far less than CNPM. ANPA's members in fact are the strongest competitors of the Postal Service. See Joint Appendix at 937. Apparently, "preprint volume in U.S. daily newspapers overall continues to grow." Id. at 935. Consequently, the PRC concluded that ANPA's competition-related arguments, like CNPM's, did not warrant raising the BRR rates. PRC Opinion at 523.

In short, the PRC adequately considered the arguments of Postal Service competitors, and addressed subsections (b)(4) and (b)(5). In evaluating competition-related arguments under subsection (b)(4), it must be remembered that the PRC's task is to protect competition, not particular competitors. Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477, 488, 97 S. Ct. 690, 697, 50 L. Ed. 2d 701 (1977). The PRC adequately did so. See also PRC Opinion at 295-96.

These subsections may be dealt with briefly; no challenges to the PRC's handling of them with respect to BRR mail are mounted in this appeal. Subsection (b)(6) requires the PRC to consider the degree of mailer preparation of the material; this it did by passing through some 80 percent of the costs avoided by handling presorted mail as compared with mail requiring sorting by the Postal Service. PRC Opinion at 524. Subsection (b)(7) concerns simplicity of rate structure; the PRC concluded that the BRR rate structure was sufficiently clear that the relatively more sophisticated BRR mailers could understand and use it. Id. Finally, it is generally agreed, and not in dispute on this appeal, that subsection (b)(8), concerning the educational, cultural, scientific and informational value of the mailed matter, is not implicated in ratemaking for BRR mail.

Finally, we address two challenges by ANPA to the BRR cost coverage and resulting rates. The first involves a possible error in the computer program used by the PRC to calculate BRR mail volume, which may have resulted in an overestimation of volume during the test year of some 5 percent. If such an error in fact occurred, revenue estimates for the test year would have been overstated by some $68 million. See Governors Decision at 4 n. 2.

] In the first place, it is by no means clear that there was any error. Assuming, however, that there was an error in computation, we hold that the Governors properly exercised their discretion in dealing with the error. As the Postal Service points out, such an error would mean that there were fewer pieces of mail over which to spread the institutional cost burden. Had the Governors determined that the resulting revenue deficiency would be too large for the Postal Service to absorb, it is likely that the entire case would have been remanded for a redetermination of the assignment of institutional costs, which might or might not have resulted in higher BRR rates. Instead, the Board determined that a revenue deficiency of $68 million was absorbable by the Postal Service. Even if all of the revenue shortfall were recouped from BRR, the Postal Service's analysis shows that the impact upon BRR rates would be de minimis. That is, if one divides the $68 million shortfall by the number of pieces of BRR mail that would result in the absence of the alleged error, namely 41.026104 billion pieces, see ANPA Brief at Exhibit 1, this would result in a required per-piece adjustment of 0.166 cents. The average revenue per piece for BRR is 10.583 cents. PRC Opinion app. G. Thus, the potential adjustment to reflect the error would amount to only 1.57 percent of the average revenue per piece. In view of the numerous uncertainties in the ratemaking process, it was not an abuse of discretion for the Governors to rule that a potential adjustment of less than 2 percent in the BRR rates did not warrant a remand and recalculation of BRR rates.

Finally, ANPA argues that the PRC abused its discretion by striking the proposal of ANPA witness Chown regarding separate rates for detached labels and their impact in general upon existing BRR rates. We reject this view. The PRC declined to consider the ANPA proposal, and the related Advo-Systems counter-proposal, chiefly because of concern over the delay that might result from the consideration of a proposal for a separate rate in this omnibus rate proceeding, which had to be completed within ten months. ANPA in its reply brief attempts to convince us that no significant additional time would have been consumed by consideration of its proposal; nevertheless, it seems to us that the PRC properly exercised its discretion when it ruled that it lacked the time to do so. "No principle of administrative law is more firmly established than that of agency control of its own calendar." City of San Antonio v. Civil Aeronautics Board, 374 F.2d 326, 329 (D.C.Cir.1967). ANPA's argument that the PRC failed to comply with subsection (b)(3), because its costing systems failed to attribute the costs incurred by detached labels to BRR mail, is flatly contradicted by the testimony of ANPA's own witness Chown, who stated that "[t]he USPS costing systems 'pick up' the costs of the label and assign these costs to third-class bulk mail." Joint Appendix at 616. To the extent that ANPA's arguments were directed at treatment of detached labels as separate pieces of mail, they were excluded from the omnibus rate case, as noted above, and to the extent that they were directed at consideration of the impact of detached labels upon BRR rates in general, they were without merit, as indicated by the testimony of ANPA's own witness Chown. We are not persuaded that Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 103 S. Ct. 2856, 77 L. Ed. 2d 443 (1983), and National Ass'n of Regulatory Util. Comm'rs v. FCC, 737 F.2d 1095 (D.C.Cir.1984), cert. denied, --- U.S. ----, 105 S. Ct. 1225, 84 L. Ed. 2d 364 (1985), are to the contrary. ANPA's arguments were made for the first time in a brief to the PRC, and were not based upon record evidence. ANPA's claim that there was no other way for it to proceed is without merit; the PRC Chairman ruled early in the proceedings that the portions of Chown's testimony pertaining to the consideration of detached labels and their impact on BRR rates in general would remain in the record, while the portions relating to a separate mail class were stricken. See Joint Appendix at 652-53. Moreover, the Chairman noted that "[i]f you have any difficulty with that, ... a motion for relief is always accepted by [this] bench." Id. In our view, ANPA's arguments relating to detached labels are without merit.

In sum, we conclude that the PRC adequately considered the record evidence and the statutory factors with respect to BRR mail, and did not abuse its discretion in determining the appropriate cost coverages or rates for that class.

The Alliance of Nonprofit Mailers (ANM) challenges the discount set for presorted third class nonprofit mail. Nonprofit organizations rely on third class nonprofit mail as the most economical way to mail brochures and circulars as well as to solicit funds. All third class mail receives some presorting, however, nonprofit mailers can presort beyond this basic level and receive a third class nonprofit discount.

The PRC set the discount for third class nonprofit mail at 80 percent of the cost savings attributable to nonprofit mailer's presorting; ANM contends that this pass through rate is arbitrary and capricious because it should be equal to 100 percent of the Postal Service's savings from handling this presorted mail. To support this contention, ANM introduced the testimony of witness Haldi, an economist, who testified at the PRC hearings that a discount of less than 100 percent violated fundamental economic principles, and, by setting the discount at less than 100 percent of attributable costs, the Postal Service in effect would be providing a subsidy to those who do not presort their mail. According to Dr. Haldi:

Third class nonprofit rates must, on average, equal attributable costs. If some nonprofit rates are set above attributable cost, other nonprofit rates must be set below attributable cost. This is precisely what happens when presort discounts are set below 100 percent of cost savings: less presorted mail pays rates below attributable costs and more heavily presorted mail pays rates above attributable costs. As a result, third class mailers are encouraged to demand too much "less presorted" mail service and too little "more presorted" mail service.

PRC Opinion at 528-29. ANM contends that the Postal Service's explanation of the discount rate represents post hoc rationalization because the reasons for the PRC decision are not stated in the PRC Opinion recommending the discount.

The Postal Service responds to ANM's challenge by pointing out that, while the PRC opinion considered ANM's position, it found that the Postal Service recommendation of an 80 percent discount rate was more appropriate than a 100 percent rate. The argument in favor of the 80 percent as opposed to a 100 percent rate was examined in the PRC's discussion of third class bulk regular rates as well as second class mail and this discussion need not be summarized again when applied to third class nonprofit discount rates. Furthermore, the Postal Service argues, the PRC must consider more than just economic factors when setting rates. We agree with the Postal Service that the discount rate for third class nonprofit presorted mail is supported by substantial evidence and we affirm the Postal Service order with respect to this discount rate.

In examining the nonprofit presort discount, "[a] reviewing court may not overturn an agency finding simply because evidence existed supporting an alternative finding. Rather, in applying the substantial evidence standard set forth in 5 U.S.C. Sec. 706(2)(E) we must first determine that a 'reasonable mind' could not find the evidence on which the agency relied to be 'adequate' to support its conclusion." Newsweek, 663 F.2d at 1210. We find that the PRC's reason in maintaining a conservative 80 percent passthrough rate is a reasonable one which can be "clearly discerned," from the record hearings, see Bowman, 419 U.S. at 286, 95 S.Ct. at 442.

The PRC considered the issue of setting presort discount levels at 100 percent of avoided costs in its opinion. When discussing the presort discount for second class mail the PRC concluded that:

We will continue to follow our conservative approach, with which the Governors agree, to set presort discounts at a level lower than the cost differentials found in Postal Service special studies. In taking this conservative approach, we do not reject the testimonies of witnesses Dalton and Baumol concerning economic inefficiency. In a perfect world where the Commission was certain that the cost differentials derived from special Postal Service studies were completely accurate, a 100 percent reflection of those differentials in proposed discounts would be acceptable. We do not live in a perfect world, and the Commission's conservative approach by not applying a 100 percent passthrough is justified on this record.

PRC Opinion at 432. Similarly, the presort differentials for third class mail were set at 80 percent of avoided costs, to " 'continue the conservative relationship between presort cost saving and presort rate discounts that prevailed in the past.' " Id. at 511 (quoting Postal Service witness Allen). The purpose of maintaining this conservative relationship is spelled out by the PRC Opinion:

[I]f the cost differential is overstated and a discount is equal to 100 percent of the differential, economic inefficiency would occur. Moreover, in that case the Postal Service as a whole would receive too little revenue. Therefore, all mail users would be harmed by those overstated rate discounts. On the other hand, with a conservative approach postal revenues are protected, and all mailers are not harmed by the presort discount.

Id. at 432.

We agree with the Postal Service that each portion of the PRC opinion should not be considered in isolation. An argument clearly applicable to several issues need not be repeated each time it applies. Were repetition required, the PRC's task of completing its complex ratemaking process within ten months, see 39 U.S.C. Sec. 3624(c), may well be insurmountable. Here, the presort discount level was discussed twice; extensive testimony was taken on the issue, and the PRC Opinion reiterated the argument in favor of a 100 percent rate in its discussion of the third class nonprofit presort discount, see PRC Opinion at 527-29. Furthermore, ANM's witness sought to refute the approach of Postal Service witness Allen when advocating a 100 percent passthrough rate. In adopting witness Allen's approach for an 80 percent passthrough rate for third class bulk regular rates, the PRC Opinion quoted from witness Allen's testimony. Clearly, the PRC should not have had to repeat its analysis of the issue in the context of nonprofit presort discount rates.

The American Postal Workers Union, AFL-CIO (APWU) challenges the PRC's recommendation of a four cent discount for five-digit presort first class mail. APWU contends that the amount of the discount is not supported by substantial record evidence and that the explanation of this discount cannot be accepted by this court because the Postal Service is engaging in post hoc rationalization.

APWU asserts that the PRC employed a new cost avoidance method in determining the presort rate. It contends that this method was not revealed until the PRC's decision was released thereby violating 39 U.S.C. Sec. 3624(a), which requires that the PRC recommend a rate decision based on record evidence. See Newsweek, 663 F.2d at 1205.

The Postal Service, however, contends that Sec. 3624(a) was not violated because the discount was not based on a new methodology, but rather was distilled from several methods presented by witnesses during the PRC's hearings. We agree with the Postal Service that the discount for first class presorted mail is not arbitrary and capricious and is based on substantial record evidence.

Pursuant to 39 U.S.C. Sec. 3624(a) of the Postal Reorganization Act, the PRC must "base its decisions upon materials presented at record hearings." Newsweek, 663 F.2d at 1205. Post hoc rationalizations of agency decisions are not acceptable, Motor Vehicle, 463 U.S. at 50, 103 S.Ct. at 2870, however, we repeat that "we will uphold a decision of less than ideal clarity if the agency's path may reasonably be discerned." Bowman, 419 U.S. at 286, 95 S.Ct. at 442.

In Newsweek, this court found that the PRC had not based its decision on record evidence when productivity adjustments made by the PRC were based on two methodologies never presented at hearings. Although the PRC had permitted briefs and comments on the new methodologies, the evidentiary record had closed and no discovery or cross examination was allowed. 663 F.2d at 1205. We remanded so that the PRC could "subject its productivity adjustment rationale to the same hearing process as all other materials upon which it bases its recommended decisions." Id. In contrast is Association of Am. Publishers, 485 F.2d 768, where the PRC based its adjustment of attributable costs between parcel post and special rate fourth class mail on an average of the figures provided by two witnesses. Id. at 772-73. As the D.C. Circuit pointed out in upholding the PRC's approach:

It would, of course, be the summum bonum if we had accurate figures as to recent costs of carrying special fourth class mail. The only available figures were inaccurate, but were susceptible of rough adjustment. The Postal Service proposed one method of adjustment; the Chief Examiner, another. The Commission regarded each as helpful, but not wholly reliable. So the Commission more or less split the difference. No doubt it would have been possible to straighten out some of the errors or supposed errors of adjustment in either the Postal Service's or the Chief Examiner's calculations. And if rate-making were an exact science such a counsel of perfection would be mandatory .... In any case the rough splitting of a difference between two fairly but not wholly satisfactory rate calculations is a familiar permissible technique.

Id. at 773.

The facts presented here are more closely analogous to Association of Am. Publishers than they are to Newsweek. The PRC did not develop a new methodology to determine the presort discount, but rather modified the approaches taken by two witnesses. Postal Service witness Cowell employed a reversion approach that focused on the costs which would be incurred if mail currently presorted were to revert to nonpresorted. PRC Opinion at 356. The PRC characterized this approach as novel, id., and concluded that the result reached by this methodology resulted in a conservative recommendation to retain the presort discount at 3cents. Id. at 356, 369-72. Several reasons were cited to support the PRC's decision to depart from the conservative approach taken by the Postal Service's witness. First, although savings from associated worksharing are not limited to presorted mail, these savings nevertheless have reduced overall postal costs. Id. at 365-66. Second, the rise in presort volume has contributed to the financial stability of the Postal Service, and this should be reflected in the presort discount. Id. at 366. The Postal Service's three cents discount was thus considered to be a floor below which the discount should not fall.

The PRC determined, however, that the 4.5cents discount advocated by the CPUM witness Bentley was a ceiling beyond which the "cost saving due to presorting can not be expected to exceed." Id. at 362. The CPUM approach was based on the Postal Services' Revenue and Cost Analysis (RCA) Report which included costs unrelated to presort savings. Thus the PRC reviewed both the approach of the Postal Service's witness as well as that of CPUM's and reached a mid-ground between them:

We accept witness Cowell's analysis as an absolute floor, below which the savings cannot be expected to fall. As stated previously, witness Cowell's reversion approach is novel in postal costing.... Witness Cowell is also very conservative in the manner in which he implements his approach.... As an estimate of the relevant cost differences which does not suffer from the deficiencies just described, we have taken the RCA report and isolated the unit difference in mail processing costs between presort and single-piece First Class. The calculations can be found in Appendix F to this decision. The method shown gives a rough approximation of the difference which can be expected in the test year. The record does not contain sufficient information to develop a more precise estimate. The difference is 4.9cents.... This figure corresponds to the one identified by witness Bentley, but does not include analysis of the operations, unconnected with sortation and hence inappropriate in deriving a presort differential.

Id. at 368-69. Far from developing a new methodology and reaching a rate not based on record evidence in violation of Sec. 3624(a), see Newsweek, 663 F.2d at 1205, the PRC has determined its presort discount rate based on the evidence presented during the hearings. It is permissible "where the regulatory administrative agency did not entirely disregard two experts, but found each somewhat in error, [to take] as its own solution a point somewhere between the two expert figures. When neither of two suggested adjustments applied to inaccurate data is completely satisfactory a rate-making body may fashion its own adjustments within reasonable limits." Association of Am. Publishers, 485 F.2d at 773; see also United States v. FCC, 707 F.2d 610, 615 (D.C.Cir.1983) (agency has "authority to adopt ... those portions of each witness' testimony that it found credible"). We find that the PRC's recommendation of a first class presort discount of 4cents to be discernable from the evidentiary record upon which the recommendation is based. See Bowman, 418 U.S. at 286, 95 S. Ct. at 442.

In addition, the APWU challenges the PRC's policy decision to encourage presorting as an usurpation of the Postal Service's authority to manage the agency. However, it can be argued that it was the Postal Service's own policy to increase presortation because the Service's recommendation of 80 percent set the discount at a higher level than the cost avoided by presorting, and the Postal Service could have rejected the discount had it been in disagreement with the PRC's approach. It is also contended that the presort discount provides a "backdoor" way to provide a lower cost business mail. The discount for presorted mail takes into account the degree of preparation by the mailer, a factor required to be considered under subsection (b)(6); it is irrelevant whether the user is a business or others who presort. We therefore deny the petition to review the Postal Service's approval of the first class presort discount.

Finally we find Petitioner DMA's argument that the Governors' failure to adjust Segment 3 Processing Costs to reflect peak load costs to be without merit. Peak load costs are additional costs allegedly created by larger volumes of mail tendered to the Postal Service at certain peak times. These costs are then attributed as part of Segment 3 costs to all classes of mail. DMA asserts that these costs should be considered institutional costs to be attributed to responsible classes of mail.

DMA contends that the Postal Service's principal costing system, the In Office Cost System (IOCS), does not distinguish between cost behavior on-peak as opposed to off-peak. As a result, DMA alleges that classes of mail which have no responsibility for peak load costs, such as BRR, are thereby prejudiced and, therefore, peak load costs should be treated as institutional costs to be attributed to responsible classes of mail. DMA concludes that the Governors' approval of the PRC recommendation of Segment 3 Processing Costs is based on inadequate data and is therefore arbitrary and capricious.

The PRC considered the issue of peak load costs extensively in its decision. Testimony was given to support DMA's contention that peak load costs should be shifted to preferential mail classes. PRC Opinion at 172-75 (testimony of ANPA witness Haldi). The PRC rejected this proposal finding that there was insufficient evidence to conclude when peak times occurred, id. at 184, and how labor patterns would be affected, id. at 185-91. DMA asserts that a Postal Service system which distinguishes between peak and non-peak costs, the Mail Processing Cost Model (MPCM), was improperly stricken from the record because the Postal Service refused to supply sufficient information to support the use of MPCM. The exclusion of this evidence, however, did not affect the consideration of peak load costs because the PRC concluded that independent reasons existed for not adopting the peak load cost proposals. See id. at 200. DMA's contention that the Board's treatment of Segment Three Processing Costs was arbitrary and capricious is thus without merit.

We hold that the Governors' approval of the schedule of postal rates and fees recommended by the PRC in Docket No. 84-1 is supported by substantial evidence and consistent with the policies underlying the Act. Accordingly, we deny the petition to review.