This case is the fourth of a series of appeals from cellular telephone licensing decisions of the FCC. The setting before us today is the Atlanta, Georgia market. The FCC awarded the Atlanta license to Gencom Cellular of Atlanta (Gen-Cell),1 to the dismay of the challengers now before us. Many of the arguments advanced by the appellant, Celcom Communications Corporation of Georgia (Celcom), and the intervenor, Cellular Mobile Systems of Georgia (CMS), were raised by their corporate affiliates in prior proceedings before this court. To the extent that such issues were resolved in our opinions in the Pittsburgh2 and Chicago3 proceedings, they will not be revisited here.

* The principal legal challenge maintained by Celcom relates to the FCC's comparative determination of geographic and population coverage. Celcom's criticism attacks the FCC's decision on two levels. First, Celcom argues that subsequent pronouncements of the FCC have wholly undermined any asserted public interest which the "coverage" criteria may once have embodied. Second, and more narrowly, Celcom attacks the FCC's use of contour coverage as the yardstick for its comparative measurement of the applicants' geographic and population coverage. After full consideration, we conclude that the preference awarded Gen-Cell for coverage is supported by substantial evidence and is neither arbitrary nor capricious.

The FCC has repeatedly affirmed that geographic and population coverage is a "major basis of comparison" of cellular applications.4 Gen-Cell, the victor in this license proceeding, had beyond cavil the greatest area and population coverage as measured by the FCC. Celcom does not attack the substantiality of the evidence supporting that preference. Nor could it reasonably do so, inasmuch as Gen-Cell's proposal indisputably covered 90,726 more people and 330 more square miles than Celcom. Celcom Communications of Georgia, Inc., FCC 85-15, at 5 (Jan. 29, 1985) (Atlanta Final Decision, hereinafter Atlanta ). Instead, Celcom argues that the differences in relative coverage relied upon by the FCC do not reflect "substantial public need." Brief at 41-42. In a supplemental brief, Celcom adds that it was not on notice that the standard of comparison was what it perceives as a simplistic rule of "more coverage is better." Supplemental Brief at 10.

We cannot agree in either respect. The FCC concluded, reasonably, that Gen-Cell's coverage was not only quantitatively greater but that its pockets reflected substantial need. Gen-Cell's pocket superiority was strengthened by its inclusion of Lake Lanier; it is undisputed that that facility draws over 16 million visitors annually. Atlanta, at 6 n. 8. The FCC found that a popular recreational area is "likely to have high mobile usage characteristics," a "significant factor" under Cellular Rulemaking. 86 F.C.C.2d at 502. Crediting service to Lake Lanier is, in our view, no different conceptually than crediting service to an airport or a major highway. Moreover, the difference in coverage offered by the respective parties is greater than that in Chicago. At 226-27. It was reasonable for the FCC to conclude, on this record, that Gen-Cell's coverage merited comparative credit.

Celcom contends that the comparison of coverage bears no relation to the public interest, not simply because such differences are "relatively minor," an argument addressed in Chicago, at 225-26, but because recent pronouncements by the FCC herald the deregulation, as it were, of cellular coverage. See Public Notice No. CL-175 (Nov. 23, 1984). According to Celcom, a cellular licensee is now at liberty to alter its contour coverage or its cellular geographic service area (CGSA) so long as the SMSA is not exceeded. At oral argument, Celcom denied that it was mounting an attack on the original comparative criteria; to the contrary, Celcom claimed that recent statements emanating from the Commission itself sapped the comparative criteria of any meaningful public-interest content.

We are persuaded that Celcom's creative analysis, when fairly viewed, mischaracterizes the FCC's position. The recent Public Notice, which describes what changes to a cellular system are "major," "minor," and "permissive," simply tracks the language of Cellular Rulemaking. Compare Public Notice No. CL-175 (Nov. 23, 1984) with 86 F.C.C.2d at 509-10. In its emphasis on flexibility, the original rulemaking provided that any applications proposing to alter CGSA boundaries would be "major" applications; that alterations in transmitter locations would be "minor" and thus would require prior approval in part to ensure FAA clearance;5 and, finally, that changes in frequency utilization would be "permissive." As we see it, Celcom's criticism in this respect ultimately reduces to an attack on the comparative criterion itself. Celcom has had an opportunity to advance this argument on numerous occasions: in the Cellular Rulemaking itself; in the several Reconsiderations and on appeal to this court from that rulemaking; on appeal to the Commission of the Designation Order in this proceeding; or in a motion before the agency to expand the issues.6 Celcom failed to do so. It is simply too late in the day to attack the comparative process itself or a comparative criterion which went unchallenged at the time it was adopted.

Celcom attempts to reinforce its argument with references to the Commission's later adoption of a lottery to replace the comparative process in markets below the top thirty. Celcom characterizes the FCC's explanations as, in effect, the confessions of a penitent regulatory agency which has belatedly recognized the public-interest bankruptcy of the comparative process. Taken in context, however, the FCC's language reveals that the lottery decision was reached not because the comparative process was deemed incapable of identifying the best applicant in the larger markets but because the difficulty, expense, delay, and administrative burden attendant to comparative proceedings had become too high. Cellular Lottery Selection, 56 Rad.Reg.2d (P & F) 8, 13-21 (1984).

We also reject Celcom's attack on the comparative measurement of contour coverage--as opposed to CGSA coverage--as the requisite point of reference. The FCC's comparison of contour coverage is not, as Celcom would have it, contrary to the criterion enunciated in Cellular Rulemaking. The major basis of comparison set forth there was "the geographic area that an applicant proposes to serve," or the "proposed service area." 86 F.C.C.2d at 502 (emphasis added). The very regulation which provides that the CGSA "shall be defined by the applicant as the area intended to be served," 47 C.F.R. Sec. 22.903(a) (1982) (emphasis added), goes on to state that the CGSA has two limited purposes, namely determining mutual exclusivity as to other systems and establishing standing for those who assert adverse effects. Id. Sec. 22.903(b). Nothing in the rules requires that the CGSA be the basis of comparison of relative coverage.

The use of contours, moreover, is reasonable. Contours measure the actual coverage which the applicant will provide with the system as designed, while the CGSA represents not the area which the applicant proposes to serve, but the potential boundaries of the applicant's license.7 What is more, both traditional FCC practice and the cellular rulemaking indicated that contours would be the likely basis of comparison. Traditionally in radio comparative proceedings, including domestic public land mobile radio proceedings, coverage has been compared by examining the "reliable service area" of the applications, as measured by the field strength contours of the stations. See, e.g., 47 C.F.R. Sec. 21.504 (1958) (currently codified at 47 C.F.R. Sec. 22.504 (1985)). In the cellular rulemaking NOPR, the FCC proposed to employ the traditional land mobile comparative criteria, such as area and population coverage, for cellular radio as well. Cellular Communications Systems, 78 F.C.C.2d 984, 1000 (1980) (NOPR). This criterion, among others, was adopted; in its cellular rules, the FCC continued to measure "reliable service area" according to the traditional contour measure. 47 C.F.R. Sec. 22.903(c) (1982). Thus, the same rule that defines CGSA also requires that the 39 dBu contours be shown, id. Sec. 22.903(a); that the Carey method be used to depict the contours "for the purpose of establishing the reliable service area of a station," id. Sec. 22.903(c); and that the applicant demonstrate that the combined 39 dBu contours will cover at least 75% of the CGSA. Id. Sec. 22.903(a).

In addition, Celcom's own application indicates that it expected that the parties' respective 39 dBu contours would be compared. The first page of Celcom's engineering statement includes a "Discussion of Coverage and Cellular Geographic Service Area (CGSA)," and discusses the "39 dBu coverage for the cellular plan" in relation to the CGSA. Celcom Engineering Statement, at 1 (June 1982). The accompanying exhibit, Map 1, identifies the 39 dBu contour area as the "Area of Proposed Coverage," and contrasts it against the "Area of Proposed CGSA" in demonstrating that the 75% requirement had been met.

Celcom raises several additional arguments in this respect. The first is that contour coverage differences are merely design differences which should be ignored, by virtue of the fact that licensees will be able to modify their proposed designs to meet the exigencies of implementation. This broadside ignores the specific context--cellular design and cell spacing--in which the FCC has declared its intention not to dictate design choices. Celcom's design-difference argument, when advanced in the coverage context, leads to the extraordinary proposition that all issues are, in the end, merely design differences, precluding any effective comparison.8 This proves far too much. Coverage is not a mere matter of design; it is the very goal of the systems being planned. Coverage is explicitly stated by the rules to be a major comparative criterion. The undeniable fact remains that Celcom proposed a smaller system than Gen-Cell, indeed, a system so much smaller that it would fail to serve over 90,000 people covered under Gen-Cell's proposal.

Celcom next claims that it was unfairly penalized by the FCC's coverage rule which, as Celcom sees it, awards aggression over caution. The pertinent FCC rule permits only de minimis extensions of CGSA beyond SMSA boundaries. 47 C.F.R. Sec. 22.903(a) (1982); see Pittsburgh, slip op. at 10. The rule exists because of the obvious fact that radio signals do not respect political boundaries; de minimis extensions are often needed to ensure full coverage of a desired portion of the SMSA. In our view, Celcom had adequate notice of the criteria which would govern the FCC's evaluation of such extensions. See Public Notice: Cellular Application Filing Procedures, at 1-2 (Mar. 24, 1982). Contrary to Celcom's claim, Gen-Cell was credited only for the area it covered within the SMSA; the de minimis extensions were not included. Celcom does not argue that the rule was enforced in an arbitrary fashion, but only that, in attempting to comply with the rule, it chose not to gamble and in the end lost. Celcom must live with its strategic choices.

Celcom's second major challenge is to the FCC's award of a slight preference to Gen-Cell for its relative demand determination. Celcom claims that the FCC erred in rejecting the ALJ's finding that no substantial evidence justified such a preference.9

The FCC rejected the ALJ's criticism of Gen-Cell's sample source (the Yellow Pages) and low positive response rate, finding instead that Gen-Cell's sample source and methodology were reasonable. Atlanta, at 10.10 Gen-Cell had "fully explained" its assumptions and methodology, see Initial Decision, FCC 84D-13, at 12-13 (Feb. 27, 1984); Gen-Cell Direct Exh.L, at 6-9, and the FCC found no serious methodological errors. Gen-Cell's survey was simple and straightforward, yielding demand data at different proposed prices. Gen-Cell Direct Exh. L, at 23.

The ALJ's primary criticism was that Gen-Cell had failed to demonstrate a nexus between its demand survey and its projection and distribution of demand; the ALJ observed that the growth rate employed in the demand projections could not be derived from the survey. Initial Decision, at 13, 29. The Commission, on the other hand, concluded that "any market demand study must contain certain assumptions which are not susceptible to proof in order to account for all relevant factors." Atlanta, at 10; Pittsburgh, at 207 & n. 53. The FCC went on to conclude that Gen-Cell's demand survey was indeed utilized to distribute demand throughout its system. These conclusions, we are persuaded, are adequately supported by the record. Gen-Cell's demand-survey data, Gen-Cell Direct Exh. L, at 24, served as a launching point for its demand forecasts, Gen-Cell Direct Exh. G, at 2, 6, which relied as well upon other growth-related assumptions and demographic data. Gen-Cell Direct Exh. G. All important assumptions were explained, rather than left to the Commission's speculation. Atlanta, at 10; Initial Decision, at 13; Gen-Cell Exhs. G, K, L.11

Moreover, serious inadequacies plagued both Celcom's and CMS's demand studies. The ALJ characterized Celcom's demand analysis as "patently defective," Initial Decision, at 28; to its credit, Celcom readily concedes that its survey was intended neither to be statistically valid nor to serve as the basis of its demand forecast. Id. at 28-29. As to CMS, Celcom's proposed partner, the demand analysis it utilized was found wanting in Pittsburgh, at 206-07 and Chicago. 96 F.C.C.2d at 1179-81. As in those applications, CMS's demand forecast for Atlanta, based as it was upon a model, was entirely unrelated to any of its surveys of demand in Atlanta. Given those defects in the competitors' studies, it was reasonable for the FCC to determine that Gen-Cell's survey, although not ideal, was comparatively better by a slight margin.

In its final line of attack, Celcom raises a trio of claims of abuse of discretion on the part of the ALJ and the Commission. We find each of them unavailing.

First, Celcom argues that the ALJ abused his discretion in denying Celcom's request for cross-examination of Gen-Cell's officials on the efficacy of the new Gen-Cell partnership proposal. Under the Commission's rules, the "[d]etermination of what, if any, cross-examination is necessary is within the sound judicial discretion of the [ALJ]." 47 C.F.R. Sec. 22.916(b)(6) (1985). The pertinent standards for exercise of this discretion under the cellular rules, as well as the APA, have been amply elaborated in Pittsburgh. At 197-99; see generally American Public Gas Association v. Federal Power Commission, 498 F.2d 718, 723 (D.C.Cir.1974); 5 U.S.C. 556(d) (1982); 47 C.F.R. Sec. 22.916(b)(6) (1985). Suffice it to say that Celcom's cursory assertions that "[t]hese are matters within Gencom's knowledge, and written testimony may not be a fully effective substitute for cross-examination," Objections to Revised Direct Case, at 3 (Sept. 26, 1983), fail to satisfy the more exacting standard imposing the burden on the requesting party to demonstrate the need for cross-examination.

Celcom next contends that the ALJ improperly permitted Gen-Cell to upgrade its application by amending it to include approximately eight lines of text describing Gen-Cell's plan for assigning signalling channels. See Gen-Cell Motion for Leave to Amend Application, at 9-10 (Mar. 4, 1983). Under the rules, major amendments proffered after the Designation Order are permitted only by leave of the ALJ upon demonstration of good cause. 47 C.F.R. Sec. 22.23(b) (1982); 47 C.F.R. Sec. 22.918(b)(3) (1982) (recodified at 47 C.F.R. Sec. 22.918(c)(4) (1985)); see Cellular Communications Systems, 89 F.C.C.2d 48, 91 (1982) (Order on Reconsideration). The ALJ found that Gen-Cell's claim of inadvertent omission flowing out of a clerical error, Motion for Leave to Amend, at 9-12 & Exh. 10, constituted good cause for acceptance of the amendment. Order, FCC 83M-995, at 2 (Mar. 28, 1983).

Celcom argues that, notwithstanding Gen-Cell's disclaimer of comparative credit, id. at 2 & n. 3, the very acceptance of an amendment rendering operable a theretofore nonfunctioning system had the forbidden effect of upgrading Gen-Cell's proposal. Both the ALJ and the Commission properly rejected this contention, explaining that the evil of "one-upmanship"12 was not implicated and that no comparative credit had been awarded for that aspect of Gen-Cell's plan. "Since the amendment merely corrected an error to properly express what Gen-Cell's plan had always been, no comparative upgrading is involved." Initial Decision, at 16 n. 32; Atlanta, at 21.

Celcom concludes its final line of attack with the claim that the Commission abused its discretion in refusing to permit a last-minute amendment of Celcom's application to reflect a stock agreement between Celcom and CMS. We disagree. The Commission's rules, as we have just seen, provide that such amendments may be filed only with leave of the ALJ upon a showing of good cause by the party seeking the amendment. Under the rules, the Commission could reasonably conclude that it would be in the public interest to accept such an amendment if (1) it were not untimely; (2) "good cause" to amend were shown; and (3) the new entity (Celcom/CMS) would have been able to seek comparative credit for a joint application which better served the public interest than the original applications. Under the circumstances before us, the findings of the Common Carrier Bureau and the Commission as to the lack of "good cause" were not unreasonable. Atlanta, at 19-20 & nn. 20-21; Order, Mimeo 4994, CC Docket No. 83-35, at 1-2 (Common Carrier Bureau June 22, 1984). This partial settlement was proposed not only after the ALJ's Initial Decision had been rendered but after all parties had filed their Exceptions to the decision. It was, moreover, a proposed settlement between the two losing Atlanta affiliates; according to the parties, the only result of the settlement would be that CMS would drop its own appeal. Finally, Celcom/CMS expressly disclaimed any comparative credit arising out of the partial settlement, Application for Review, at 2, 4-5 (July 13, 1984); in any event, such credit is, as we have already observed, precluded by the "one-upmanship" prohibition of the rules.

As we read the record, the Commission's action had no impact on the partial settlement itself, which may proceed with or without FCC sanction; the Commission concluded only that "good cause" had not been shown to permit a losing party to amend its application to reflect a stock agreement. This action scarcely rises to the level of a reversible abuse of discretion. The FCC reasonably found that a belated amendment of this sort did nothing to advance the goals of the cellular settlement policy but would, to the contrary, spawn yet further delays in a delay-ridden process.

In summary, we conclude that the comparison of geographic and population coverage within the parties' respective 39 dBu contours, rather than their CGSA's, does not contravene the dictates of Cellular Rulemaking, and that the Commission could reasonably find that such a comparison adequately serves the public interest. We adhere to our holding in Pittsburgh that the parties to these comparative proceedings had adequate notice that their relative demand determination methodologies would be subject to comparative evaluation. Finally, we hold that the acceptance of one amendment upon a showing of good cause, rejection of another due to the absence of such a showing, and the denial of an inadequately supported request for cross-examination, did not work an abuse of discretion. We therefore conclude that the FCC's ultimate award in the Atlanta cellular market reflects reasoned decisionmaking which is adequately supported by the record.

Affirmed.