Cellular Mobile Systems of Illinois, Inc. v. Federal Communications Commission, Rogers Radiocall, Inc., Intervenor

Opinion for the Court filed by Circuit Judge STARR. STARR, Circuit Judge: This case and its companion case, Cellular Mobile Systems of Pennsylvania, Inc. v. FCC, 1 decided on December 20, 1985, present highly similar issues for resolution. Cellular Mobile Systems of Illinois, Inc. (CMS) is a disappointed applicant for a nonwireline cellular license in the Chicago market. The claims here essentially mirror those in the Pittsburgh appeal. We conclude that CMS’s contentions fare no better than those of its sister applicant in Pittsburgh and accordingly affirm the Commission’s decision. I CMS and its sole nonwireline competitor in the Chicago market, Rogers Radiocall, Inc. (“Rogers”), filed their respective applications and direct case exhibits on June 7, 1982, as did the applicants in the top thirty markets across the Nation. In November 1982, the FCC issued its first Designation Order, 2 holding that both Chicago nonwireline applicants satisfied the basic qualifications and setting the two applications for comparative hearing on what were destined to become the four standard cellular comparative issues. 3 These issues reflect the comparative criteria laid out in the FCC’s cellular rulemaking. 4 After the AU in his two initial orders mandated the procedural format to be followed, CMS claimed a “right” to surrebuttal, and made a generalized request for discovery. CMS admitted that the AU had authority to permit discovery only if the Designation Order revealed a need for underlying data, yet failed to point out any specific need for data which might have been indicated in the Order. Request for Prehearing Conference, at 11 (Nov. 19, 1982). CMS’s two requests were denied. Rebuttal cases were then filed in December 1982. The first hearing session was held two months later in February 1983. The initial session was limited to receipt of direct and rebuttal exhibits, ruling on objections to exhibits, and ruling on requests for cross-examination. See 47 C.F.R. § 22.-916(b)(5) (1984). Citing the Cellular Rule-making standard for permitting cross-examination, the AU permitted cross-examination by Rogers of one CMS witness concerning CMS’s system design and ability to expand to accommodate demand. Transcript at 383-581, 606-610. CMS, in turn, was permitted to cross-examine two Rogers witnesses, one regarding the relationship between paging service and cellular service, id. at 157-71, 181-83, and the second *217 with respect to (1) system expansion and (2) two rebuttal exhibits which treated CMS’s expansion plans. Id. at 185-353, 368-372. CMS subsequently withdrew its request to cross-examine the second witness, electing to rely instead on its rebuttal case. Id. at 104-105. 5 On the other hand, CMS was denied cross-examination which it sought of four direct witnesses and eight rebuttal witnesses. Cross-examination was thereafter conducted over a three-day hearing period in April 1983. The Separated Trial Staff sought surrebuttal, which was denied, id. at 90-91, 101-102, a ruling to which CMS never objected. At the fourth day of hearings, the record was closed. Id. at 611-16. Proposed Findings of Fact and Conclusions of Law and Replies were filed in May 1983. The AU issued his. Initial Decision on August 17, 1983, recommending that Rogers’ application be granted. 6 Rogers received a substantial comparative preference under Issue (a), see note 3 supra, based on its proposed coverage of geographic area and population, its superior demand projections, and its ability to accommodate anticipated local demand. Rogers also received a substantial preference under Issue (b) for its superior expansion plan, and two slight preferences under Issue (c) for its maintenance and personnel proposals. Overall, these preferences outweighed CMS’s slight preference under Issue (a) for ability to accommodate anticipated demand for roamer service, and its slight preference under Issue (c) for its proposed rate structure. Initial Decision, 96 F.C.C.2d at 1225. Under the first sub-issue of Issue (a), geographic area and population coverage, the AU found that Rogers’ proposed service area was 127 square miles greater than that of CMS and would be able to serve 58,000 more people. Id. at 1218. 7 Each proposal included five discrete “pockets” not covered by its competitor. The pockets covered by Rogers but excluded by CMS had a population of 111,781 and were estimated to have 72 projected cellular users. CMS’s exclusive pockets had a population of 67,484 and were estimated to generate 42 users. Id. at 1197. The AU found that Rogers’ pockets generated more potential users than CMS, inasmuch as the former’s areas were primarily more densely populated municipal service areas and rural centers, containing more areas of special user concentration. Id. Although the AU found that CMS would have provided greater “depth of coverage,” id., 8 he declined to award a preference for this feature, principally because this factor was not encompassed by the designated issues. Id. at 1218. What is more, the AU criticized CMS’s “depth of coverage” as “both deceptive and meaningless” because, as CMS admitted, its comparative strength, if any, was premised on the applicants’ initial systems and would be rendered, an “extremely shortlived” superiority during system expansion. Id. at 1218-19. Moreover, CMS failed to show, the AU concluded, that the cell overlap contemplated by Rogers’ proposal would be insufficient. Id. at 1219. The AU also found that CMS’s multiple cell overlap would generate excessive co-channel interference. Id. Overall, Rogers’ proposal was preferred in this respect. The AU also preferred Rogers’ proposal under the second sub-issue of Issue (a), *218 namely determination of demand. 9 Rogers’ demand forecast was derived from a demand survey of 469 randomly selected businesses in Chicago. The survey produced data on consumer interest at various costs, as well as usage patterns. Id. at 1198. Compucon, a marketing firm retained by Rogers, employed the data to forecast three demand “scenarios” — expected demand, optimistic, and pessimistic — which reflected adjustment of different variables, such as rate of business growth, cost of service, and market share. Id. at 1199. For each “expected” demand projection, Rogers stated its assumed rate of business growth, expected lease cost, expected number of units per customer, assumed rate of penetration of market demand, and expected market share. Id. To locate cellular demand, Compucon then collected demographic data by census tract on key variables, such as construction activity and expected traffic flow. This data was then used to identify high potential demand areas. Within these areas, census tracts were ranked according to demand-influencing variables such as population, the number of businesses, and income; demand was distributed accordingly among census tracts. Id. at 1199-1200. 10 CMS’s demand survey strategy was the same as that employed by its unsuccessful Pittsburgh affiliate. As will be recalled from that proceeding, the Claritas survey employed by CMS analyzed demand for paging services in a few unnamed markets in which CMS’s parent, Graphic Scanning, offered such services. This study projected demand for 67,155 pagers in Chicago. Id. at 1200. CMS’s Westat study was a telephone survey of businesses in the thirty largest SMSA’s; this survey estimated a need for 95,400 mobile units and 86,900 portable units in Chicago during the first five years. Id. at 1201-02. Four additional surveys, two of Graphic Scanning’s Chicago paging customers and two of the general business population in Chicago, also collected data to verify the demand projected. Id. at 1200-01. CMS then used the same model employed in Pittsburgh. Based on the Claritas data, the top thirty SMSA’s were grouped into three categories based upon the level of expected paging demand. Chicago was assigned a need factor of 0.0045, or 4.5 cellular units per 1,000 population, as were eight other SMSA’s. CMS then multiplied this factor by the population of Chicago to yield a projected fifth-year total market demand for cellular telephones of 31,959 units. Id. at 1202. 11 Based on all this, the AU found that, while Rogers’ demand forecast was based on certain unproven assumptions, its assumptions and methodology were nonetheless explicitly stated and appeared to be reasonable. CMS, in contrast, was criticized for failing clearly to state its operating assumptions and to explain its methodology. Like Rogers, CMS had assumed that the potential users would be primarily *219 businesses, yet CMS based its projected demand on the general population. Id. at 1219 n. 37. As in Pittsburgh, the CMS application appeared to ignore the presence of a wireline competitor, id. at 1219, and, most critically, assumed that the demand for paging service was closely related to the demand for cellular service. Id. at 1220. The AU also found that CMS’s “greatest deficiency” was its “failure to establish a nexus between the findings of its market studies and its projections of demand and demand distribution.” Id. at 1219. The AU found that CMS’s methodology raised “serious doubts” as to whether its demand projections reflected a sound, specific study of the Chicago market or were, instead, extrapolations from “national generalizations of unknown relevance” to Chicago. Id. at 1220. Rogers was also awarded a preference for its superior ability to accommodate demand, the third sub-issue under Issue (a). Rogers’ consulting engineer described how the ten-year demand forecasts, user densities per census tract, and population growth and projected community development patterns were utilized in such matters as the initial selection of cell sites and determining the channel requirements per cell needed to achieve the desired grade of service. Id. at 1203-04. CMS’s proposal, like that of its affiliate in Pittsburgh, configured an initial system capable of accommodating fifth-year demand. The AU found that CMS’s switching equipment was insufficient to handle the numbers of channels proposed. Id. at 1205, 1221. The AU also concluded that Rogers’ system would always provide more channels than CMS to serve the same amount of traffic, thus providing a superior grade of service. Id. at 1221. In contrast to the detailed descriptions by Rogers’ expert, CMS’s expert merely asserted without explanation that CMS had relied on demographic and demand data to design its system. However, the AU found this claim to be empirically unsupported. Id. at 1205-06, 1220; Transcript at 438-40. The CMS model, moreover, presumed a nucleus of high user-density cells, all of which were assigned the same number of channels, and which were surrounded by a ring of medium user-density cells, which were, in turn, surrounded by a ring of low user-density cells with, again, all the same number of channels. 96 F.C.C.2d at 1206. Within these various groupings, cells might have quite different populations, yet would have the same number of allocated channels. Id. at 1206 & n. 28. As did Judge Ehrig in the Pittsburgh proceeding, AU Fitzpatrick determined that CMS’s system would potentially be afflicted with co-channel interference and degraded service. Id. at 1206, 1220. 12 Rogers also received a substantial preference under the second major issue, Issue (b), for its superior proposal for expanding system capacity in a coordinated manner. The AU found that CMS, contrary to its claim, had not tailored its expansion plan to the expected growth in demand in Chicago, but had employed the same standard growth formula used in all markets. Id. at 1222. CMS was criticized for configuring its system to meet five-year demand, inasmuch as its expansion plan would have to be substantially modified if demand were to develop in a manner not predicted by CMS. Rogers’ “phased-in” expansion plan, in contrast, was deemed more flexible. Id. CMS was also criticized for inadequate future switching capacity and undue intrasystem interference. Id. Rogers, in contrast, was praised by the AU for higher proposed quality of service beyond the first year and greater capability for reuse of frequencies. Under Issue (c), the least weighty of the comparative factors, Cellular Rulemaking, 86 F.C.C.2d at 503, CMS was awarded a slight preference for its proposed rate structure which better reflected both a subscriber’s actual volume of usage and pro *220 gressive time-of-day pricing. 96 F.C.C.2d at 1223. Rogers was awarded one slight preference for its more detailed maintenance proposal and a second slight preference for its personnel plan. Consequently, Rogers was preferred overall on Issue (c). The Commission released its Final Decision 13 in March 1984 affirming the AU’s grant of Rogers’ application. As in the Pittsburgh decision, the Commission reduced or eliminated many of the preferences awarded by the AU to the successful applicant and reversed many of the adverse findings with respect to CMS’s system design and expansion plan. The net result was that Rogers’ substantial preferences under Issues (a) and (b) were reduced to slight preferences, and both of its Issue (c) slight preferences were eliminated altogether. However, Rogers’ slight preferences under Issues (a) and (b) continued to outweigh CMS’s retained slight preference under Issue (c). Chicago, 96 F.C.C.2d at 1192-93. The Commission found that these differences were “sufficient to make a choice under the comparative criteria previously established,” id. at 1175, concluding that the Rogers proposal was superior. 14 First, the Commission affirmed the AU’s conclusion that Rogers’ proposed geographic and population coverage was superior. 15 CMS claimed in this respect that the AU had relied exclusively on residential population statistics to compare the applicants’ respective coverages and based his conclusions on insignificant differences. Agreeing that no necessary correlation existed between residential population and cellular demand, the Commission determined that the AU had indeed examined non-common areas, or “pockets,” and affirmed the finding of a “significantly higher level of demand” as reflected by “high mobile usage characteristics” in Rogers’ pockets. Id. at 1177. As in Pittsburgh, the Commission declined to grant comparative credit for CMS’s admittedly greater “depth of coverage,” for CMS had “not demonstrated that building penetration and service to portables would be inadequate without its proposed depth of coverage.” Id. Second, the Commission awarded a preference to Rogers for its superior determination of the anticipated level of demand. Rogers had not only determined total demand by using a Chicago-specific survey and considering key forecasting variables, but had also distributed the projected demand across the market on the basis of “meticulously projected high potential usage areas.” Id. at 1181. The Commission rejected CMS’s contention that, by virtue of the fact that the two applicants’ demand *221 projections were quite similar for certain years, the AU had erred in comparing methodologies. In the Commission’s view, an applicant could design a system to accommodate demand only if it had first projected the expected demand, distributed it across the CGSA and then used that demand analysis in system design. The Commission agreed that CMS’s methodology had “serious shortcomings.” Id. CMS had failed to submit evidence to support adequately the reasonableness of its demand estimations and distribution. All six of its surveys indicated differing levels of demand; moreover, CMS had acknowledged that it relied on only two studies, the Claritas paging survey and the Westat telephone survey. Id. at 1178; CMS Proposed Findings and Conclusions, at 16-17 (May 16, 1983). The Commission concluded that CMS had failed to show how, if at all, it had employed the Westat survey or the quartet of additional Chicago-specific surveys. 96 F.C.C.2d at 1181. As in the Pittsburgh application, key assumptions underlying the Claritas survey were left unexplained. CMS failed adequately to demonstrate its fundamental assumption, namely that a sufficient link existed between paging demand and cellular demand to warrant reliance on paging data. Id. Consequently, in the Commission’s view, CMS had failed to indicate how the Claritas survey was relevant to cellular demand, much less to the Chicago market. CMS had surveyed only businesses, yet expressed its projected demand in terms of the general population. CMS’s expected demand was expressed in terms of “total market need,” thus indicating that the presence of a wireline competitor had not been factored into its analysis. Id. In contrast, Rogers’ application differed markedly in that its assumptions and methodology were both “reasonable” and “explicitly stated.” Id. The FCC, thirdly, awarded Rogers a preference for its superior ability to accom- modate local demand, inasmuch as its proposed system design, particularly its channel allocation plan, was based upon its forecast of demand distribution. Id. at 1182-83. In contrast, the FCC concluded, the record supported the ALJ’s finding that CMS had not relied on its demand distribution forecasts. CMS had designed its system using a model, but had failed to demonstrate the model’s relevance to Chicago. Id. at 1183. The Commission remained unmoved by CMS’s reprise of its “depth of coverage” design philosophy, concluding that no comparative credit was due because CMS had again “failed to demonstrate that its design [was] superior.” Id. The FCC also rejected CMS’s claim that it deserved a preference because of its superior qualifications due to the experience of its parent, Graphic Scanning Corporation. Id. at 1182 n. 27. To CMS’s chagrin, the FCC affirmed the finding of co-channel interference due to the assignment of a common signalling channel to overlapping cells, but concluded that this fact was “not decisionally significant” inasmuch as the defect was so easily remedied. Id. at 1182-83. 16 The FCC, moreover, reversed the ALJ’s findings that CMS’s switching capacity would be inadequate. Id. at 1182. With regard to Issue (b), the Commission generally affirmed the preference awarded to Rogers. CMS, however, successfully persuaded the Commission to reject much of the AU’s criticism of the CMS proposal, thus reducing Rogers’ advantage under this issue from a substantial to a slight preference. Id. at 1188. The ALJ erred, the Commission concluded, in refusing to accept CMS’s supplemental evidence with respect to switching capacity. The Commission then rejected the ALJ’s finding that Rogers proposed a superior grade of service; rejected as “irrelevant” the use of frequency reuse ratios, or degree of frequency reuse, as a basis for comparison; and reversed the finding of undue intra *222 system interference among the voice channels. Id. at 1185-87. As in Pittsburgh, the FCC rejected the ALJ’s “fundamental principle” of uniform cell spacing in cellular design. Id. at 1186. This concept, the Commission opined, violatéd its chosen path of refraining from “imposing any rigid system design concepts in order to permit applicants the flexibility (within certain parameters) to design their systems in different ways.” Id. at 1186, 1188 (citing Cellular Rulemaking, 86 F.C.C.2d at 507). The Commission determined that CMS’s design, although different, was “viable”; “[w]e will not assess CMS a demerit merely because it chooses not to follow the conventional wisdom.” Id. at 1188. However, CMS was unable to eliminate Rogers’ preference under Issue (b) entirely, nor was it able to capture a preference of its own. Describing flexibility as “the touchstone of cellular technology,” the Commission observed that the cellular carriers’ expansion plans were expected to change “as experience better enable[d] the carrier to assess the needs of the market.” Id. The Commission embraced the finding that CMS’s proposal to build an initial system to accommodate fifth-year demand was “inherently somewhat less flexible” than Rogers’ “phased-in expansion approach.” Id. CMS also succeeded in eliminating Rogers’ preferences under Issue (c). The Commission affirmed CMS’s slight preference for its proposed rate structure, which better promoted the Commission’s policy of favoring “usage-sensitive, cost-based pricing.” Id. at 1190. The Commission vacated both of the slight preferences Rogers had received for its maintenance and personnel proposals, on the ground that neither party’s proposals in these areas had been shown measurably to affect their ability to implement their respective plans. The FCC denied CMS’s Motion to Reopen the Record, which sought surrebuttal on the issues of signalling channel interference, switching capacity inadequacy, intrasystem interference, and principles of cell spacing. The Commission determined that CMS had not been prejudiced in this respect, since the FCC either had overruled the AU on those issues or had not accorded them any weight in its determination. Overall, CMS’s retained slight preference under Issue (c) was insufficient to counterbalance Rogers’ slight preferences under both Issues (a) and (b). 17 Subsequently, CMS sought reconsideration and filed a second Petition to Reopen the Record; both petitions were denied. Rogers Radiocall, Inc., 56 Rad.Reg.2d (P & F) 951 (1984) (Chicago Reconsideration Order). II On appeal, CMS raises several procedural challenges which are substantially similar to those raised in Pittsburgh. We dealt at length in that opinion with the procedural standards governing such claims and will therefore not revisit those questions here. Pittsburgh, at 197-199, 202. (CMS’s specific protests as to the denial of cross-examination of one witness and denial of surrebuttal are addressed infra.) Moreover, here, as in Pittsburgh, CMS claims that it was denied fair notice of the comparative critieria which the Commission employed. Our analysis in Pittsburgh is controlling, see Pittsburgh, at 202-203; we conclude not only that Cellular Rulemaking provided ample notice that demand determinations would be utilized in comparing the applicants’ respective proposals, but that CMS’s own application amply demonstrates that is had actual notice of this requirement. A CMS first challenges the denial of cross-examination of Mr. Goldman, a surrebuttal witness for Rogers. We find the attack unavailing. As we saw in Pittsburgh, the *223 Cellular Rulemaking placed the availability of cross-examination within the “sound judicial discretion of the judge, the prevailing standard being whether the person requesting cross-examination has persuasively demonstrated that written evidence is ineffectual to develop proof.” Cellular Communications Systems, 89 F.C.C.2d 58, 92 (1981) (Order on Reconsideration). CMS argues that, on cross-examination, it would have challenged Goldman’s criticisms of (1) CMS’s assumption that the market demand for paging service was an appropriate model for cellular demand; (2) CMS’s failure to recognize differences in cellular demand in different markets; and (3) CMS’s application of business demand analysis to the residential population. CMS Brief at 26-27 & n. 27. In its request for cross-examination, however, CMS wholly failed to describe specifically these areas as proposed subjects of cross-examination. Instead, CMS proposed in broad terms to examine Mr. Goldman regarding (1) his assessment of the relative importance of the applicants’ pocket areas; (2) his “understanding of CMS's demand forecasting methodology and his ‘critique’ of that methodology”; and (3) his assertion regarding channel capacity, in order to demonstrate his lack of understanding of CMS’s “depth of coverage” concept. CMS Objections to Rebuttal Exhibits and Requests for Cross-Examination, at 7 (Jan. 17, 1983). CMS failed to point to any specific weakness in the proof or to suggest necessary questions or otherwise to demonstrate that the written, evidence was ineffectual to develop proof. Given this failure to satisfy the appropriate standards, the denial of cross-examination was not error. Moreover, CMS fails, in our view, to demonstrate any prejudice in this respect, for it seeks not to test the evidence but only to employ cross-examination in an attempt to fill gaps in its own case. CMS argues that, through cross-examination (or surrebuttal), it would have (1) shown that the market demand survey methodology is not necessarily to be preferred to a modeling methodology in forecasting demand, and that Mr. Goldman’s firm likewise employed modeling techniques in other cellular proceedings; (2) confronted Goldman with evidence of the similarities between paging service and cellular service; (3) demonstrated that its forecast of demand volume and location reflected its market research results; and (4) demonstrated its recognition of the unique characteristics of various markets. CMS Brief at 28-29. But, as we said in Pittsburgh, cross-examination is an improper mode for raising contentions which properly belonged in CMS’s Direct and Rebuttal cases. That, at bottom, is what should have occurred here. Equally fundamental is CMS’s failure to demonstrate to our satisfaction that either the ALJ or the Commission relied upon Mr. Goldman’s rebuttal testimony; it appears to us that the basis in the record for the ALJ’s and the Commission’s factfindings and conclusion is found in CMS’s own Direct Case Exhibits, not in Rogers’ rebuttal evidence. See Chicago, 96 F.C.C.2d at 1177-89; Initial Decision, 96 F.C.C.2d at 1200-02. Furthermore, in advancing these claims, CMS itself engages in the very practice of sandbagging which it so heartily decries, for CMS never appealed in a timely fashion the AU’s denial of cross-examination of Mr. Goldman or any of these issues to the Commission. Contrary to its assertion here, CMS Brief at 27, CMS’s Exceptions, Motion to Reopen the Record, and Petition for Reconsideration are silent as to denial of cross-examination; the first two filings sought only to rebut Rogers’ evidence on the subjects of interference and uniform cell spacing, issues which were either disregarded by the Commission as being inconsequential or reversed altogether. CMS Exceptions at 4-7 (Sept. 23, 1983); CMS Motion to Reopen the Record (Sept. 23, 1983). CMS first raised the cross-examination issue long after the Commission’s final decision. Petition to Reopen the Record, at 22-23 (May 22, 1984). CMS tenders the further argument, related to its cross-examination challenge, that it was improperly denied an opportuni *224 ty for surrebuttal on these issues. As with cross-examination, surrebuttal is not available by right, but rests in the sound discretion of the ALJ. Simply stated, surrebuttal is authorized when the ALJ finds that it is needed to complete the record. Order on Reconsideration, 89 F.C.C.2d at 92; 47 C.F.R. § 22.916(b)(6) (1984). The only request to the AU was a highly generalized suggestion set forth in a motion for a prehearing conference lodged over a month before rebuttal cases were submitted. Request for Prehearing Conference, at 9-1Ó (Nov. 19,1982). Given this timing, the AU could scarcely determine if such additional testimony was needed to complete the record. 18 CMS never renewed its request for surrebuttal before the AU; instead, CMS’s first specific request for surrebuttal came in its first motion to reopen the record. As it addressed only the issue of intrasystem interference, an issue on which CMS eventually prevailed, CMS was obviously not prejudiced by its denial. - CMS’s second specific request for surrebuttal was proffered two and one-half months after, the full Commission’s final decision and one and one-half months after CMS filed its Petition for Reconsideration. Petition to Reopen the Record (May 22, 1984). Denial of this belated request was likewise proper, for none of the three proffered surrebuttal exhibits presented newly discovered evidence; CMS therefore failed to show such “unusual or compelling circumstances” as would permit an additional opportunity to present evidence simply to strengthen its case. Heart of Georgia Broadcasting Co., 19 F.C.C.2d 20, 30-31 (Rev.Bd.1969). B CMS’s second broad-based attack is that the FCC’s application of its comparative standards under Issue (a) in this case violated the requirements of reasoned decision-making and was thus arbitrary and capricious. CMS argues that the standards were inconsistently applied and resulted in a decision which was premised on trivial distinctions and ignored material differences. CMS also argues that the decision to award a preference to Rogers under Issue (b) was unsupported by substantial evidence. First, CMS claims that the inconsistent application of standards under Issue (a) renders this decision arbitrary and capricious. CMS argues that Chicago is both internally inconsistent, in that certain standards were inconsistently applied to the parties or to different issues, and inconsistent with subsequent cellular proceedings. The first point of alleged inconsistency flagged by CMS is the Commission’s statement (with reference to the entire comparative process) that it was “constrained to compare the applicants on the basis of the systems they have proposed, not on the basis of what might ultimately evolve.” 96 F.C.C.2d at 1174. CMS argues that this approach compels the Commission to compare the relative “depth of coverage” offered by the applicants. CMS Brief at 47. This argument, while cleverly couched, is unmeritorious; as we observed in Pittsburgh, CMS’s oft-repeated “depth of coverage” claims were consistently evaluated and rejected for two reasons. Not only was depth of coverage not a comparative issue in the first instance, but CMS had failed to satisfy its burden of demonstrating (1) that its much touted “depth” was necessary to satisfy one of the existing criteria and (2) that its proposal was superi- or to its competitor’s or, conversely, that its competitor’s proposal was inadequate. Pittsburgh, at 211, 212; Chicago, 96 F.C.C.2d at 1177, 1183. The second alleged inconsistency is the claim, identical to that advanced in Pitts *225 burgh, see at 206-07, that the Commission openly accepted various assumptions of CMS’s opponents, yet rejected CMS’s assumption that paging demand was indicative of cellular demand. CMS Brief at 48; see Chicago, 96 F.C.C.2d at 1181. This contention is addressed sufficiently in Pittsburgh, at 206-07. In brief, the agency's decision here, as there, indicates that Rogers’ assumptions and methodology were accepted because they were both reasonable and explicitly stated. Chicago, 96 F.C.C.2d at 1180-81. In contrast, the CMS assumption, so fundamental to its demand determination, was unsupported by evidence which could undergird a finding of reasonableness. Id. at 1181. CMS next raises the specter of decisional inconstancy between Pittsburgh, Chicago, and subsequent cellular decisions. CMS Brief at 42-45, Reply Brief at 5-19. None of these arguments sway us from the view, articulated in detail in Pittsburgh, at 207-OS, 210-11, 213-14, that this appeal, as one of the two lead proceedings, properly provides the precedential standard against which later decisions aré to be measured. It is the subsequent cellular decisions which may be subject to challenge should they lack fidelity to (or a reasoned explanation of a departure from) these lead decisions. CMS next argues that the Commission’s decision was arbitrary and capricious in that, contrary to its duty to compare and credit material differences among the applicants, the Commission instead awarded a preference for geographic and population coverage based upon inconsequential distinctions. Rogers and CMS proposed to serve .98.6% and 97.8%, respectively, of the Chicago SMSA population. Chicago, 96 F.C.C.2d at 1175. Both covered a common core area; in addition, each covered “pockets” not covered by the other. The Commission found that Rogers’ pockets would generate thirty more users than CMS’s pockets, id., a number CMS characterizes as an “infinitesimal portion of the population in the area to be served.” CMS Brief at 59. 19 In objecting to the Commission’s geographic coverage analysis, CMS, in effect, asks us not only to find the decision arbitrary, but also to find, reviewing the weight of the evidence, that the preference awarded Rogers was unsupported by substantial evidence. Although CMS may find the distinctions minor, the Commission’s decision reflects reasoned consideration of designated factors of which CMS had long been well aware: area, population, highways, population density, and likelihood of high mobile usage characteristics. See Cellular Rulemaking, 86 F.C.C.2d at 502-03. CMS acknowledges that it believed geographic coverage to be one of the two principal bases of comparison. CMS Brief at 33-34. There was absolutely no doubt that the applicant which proposed greater coverage, as measured primarily by geographic area and population, would receive the preference. 20 *226 What is more, CMS fails accurately to recount the actual differences discerned by the Commission. Rogers proposed to serve 127 more square miles and 58,000 more people than CMS. In addition, the FCC concluded that Rogers’ pocket areas would not only generate 30 more users than CMS's primarily agricultrual pockets, but also reflected greater high mobile usage characteristics. Chicago, 96 F.C.C.2d at 1175-76; Initial Decision, 96 F.C.C.2d at 1196-97. These are more than sufficient grounds to undergird a reasoned agency decision; we conclude that the substantial evidence standard is amply satisfied in this respect. CMS nonetheless protests that while Rogers’ coverage was indeed greater, mere superiority in this respect was somehow not enough. This contention flies in the face of the crystalline clarity of this criterion’s mandate to offer greater coverage in order to merit the preference. The fact that the disparity is not enormous simply does not render unreasonable a decision based upon an undisputedly modest disparity, but a disparity nonetheless. CMS then returns to its recurrent theme in these proceedings — its “depth of coverage” concept. In both Pittsburgh and the instant case, CMS repeatedly sought preferences, alternatively under the issues of “geographic coverage,” “ability to accommodate demand,” and “expansion plan,” for its superior depth of coverage. Here, CMS claims that the FCC shirked its duty to compare the applicants on all material differences, including design differences, thus violating the public interest standard. As we elaborated in Pittsburgh, at 211- 12, the FCC did not err in declining to compare cellular designs. By virtue of its policy of encouraging technical diversity, the FCC appropriately refused to hold that greater cell overlap, in itself, warranted a preference. Here, as in Pittsburgh, CMS failed to demonstrate that its depth of coverage, if it was in fact superior, improved its comparative stance under the designated issues. Finally, CMS criticizes Rogers’ receipt of a preference under Issue (b) as founded upon erroneous factfinding and thus unsupported by substantial evidence in the record. The FCC awarded a slight preference to Rogers in this respect because of the proposal’s greater flexibility; this judgment was based on the “phased-in” nature of its expansion plan, which reflected Rogers’ demand studies. 96 F.C. C.2d at 1188. CMS’s argument is premised on the claim that the FCC’s decisions reflect “two areas of decisional consequence,” flexibility and detail. CMS Brief at 50. Significantly, CMS does not challenge the FCC’s conclusion that Rogers’ expansion plan was indeed more flexible. CMS Brief at 49-53. CMS argues instead that its opponent’s expansion plan lacked sufficient detail to merit a preference, suggesting that the record is devoid of evidence of Rogers’ channel capacity and per-cell channel allocation. CMS Brief at 51-52. However, this contention was specifically addressed and rejected by the AU, who explained precisely the manner in which he was able to derive that information from the record. Initial Decision, 96 F.C.C.2d at 1208 n. 30. 21 When *227 this matter was raised before the Commission on reconsideration, it was again rejected. The Commission upheld the adequacy of the AU’s findings, stating that “CMS has presented no new evidence to persuade us that Rogers’ expansion plan lacks sufficient detail to merit a preference.” Chicago Reconsideration Order, 56 Rad.Reg.2d at 954. It cannot be gainsaid that this is a precise, considered articulation of the facts found, nor can it be doubted that the Commission’s finding in this regard is supported by substantial evidence in the record. Moreover, as we observed in Pittsburgh, when asked to review such highly technical determinations, we are “obliged to decline the invitation for this highly technical task necessitates considerable deference to the specialized agency’s expertise.” Pittsburgh, at 213; MCI Cellular Telephone Co. v. FCC, 738 F.2d 1322, 1333 (D.C.Cir.1984). Ill In this proceeding, as in the companion Pittsburgh case, we conclude that CMS had fair warning of the comparative criteria it was expected to meet. In this proceeding, unlike Pittsburgh, limited cross-examination was made available by the ALJ. We conclude that the AU did not abuse his discretion in denying either the additional cross-examination or surrebuttal sought by CMS. CMS’s various other claims also fail, for the reasons already stated. We hold that the FCC’s decision here, as in Pittsburgh, reflects reasoned decisionmaking and is supported by substantial evidence in the record. Affirmed. Footnotes [1] Cellular Mobile Systems of Pennsylvania, Inc. v. FCC, 782 F.2d 182 (D.C.Cir.1985) (Pittsburgh). [2] Advanced Mobile Phone Service, Inc., 91 F.C. C.2d 512 (1982) (Chicago Designation Order). [3] The comparative issues were: (a) to determine on a comparative basis the geographic area and population that each applicant proposes to serve; to determine and compare the relative demand for the services proposed in said areas; and to determine and compare the ability of each applicant’s cellular system to accommodate the anticipated demand for both local and roamer service; (b) to determine on a comparative basis each applicant’s proposed for expanding its system capacity in a coordinated manner within its proposed CGSA in order to meet anticipated increasing demand for local and roamer service; (c) to determine on a comparative basis the nature and extent of the service proposed by each applicant, including each applicant’s proposed rates, charges, maintenance, personnel, practices, classifications, regulations and facilities (including switching capabilities); and (d) to determine, in light of the evidence adduced under the foregoing issues, what disposition of the referenced applications would best serve the public interest, convenience and necessity. Chicago Designation Order, 91 F.C.C.2d at 520-21 (footnotes omitted); see also Pittsburgh, at 188-89. [4] Cellular Communications Systems, 86 F.C.C.2d 469, 502-03 (Report and Order), modified, 89 F.C.C.2d 58 (1981), further modified, 90 F.C.C.2d 571 (1982), petition for review dismissed sub nom. United States v. FCC, No. 82-1526 (D.C.Cir. Mar. 3, 1983) (hereinafter Cellular Rulemaking). [5] The Separated Trial Staff of the FCC’s Common Carrier Bureau was permitted to cross-examine that witness. CMS was permitted to examine him to "the extent that doors are opened.” Transcript at 113-117. [6] Rogers Radiocall, Inc., 96 F.C.C.2d 1194 (1983) (Initial Decision). [7] The AU made the following findings as to the parties' proposed service areas: Geographic Area Population Rogers 3,212 86.3% 7,005,000 98.6% CMS 3,086 82.9% 6,947,000 97.8% Initial Decision, 96 F.C.C.2d at 1196. The Chicago SMSA encompasses 3,724 square miles and a population of 7,103,624. Id. [8] “Depth of coverage” is a concept employed by CMS to describe its high degree of cell concentration, resulting in greater cell overlap. [9] The two applicants forecast demand as follows: Year Rogers CMS 1 7,120 4,802 2 9,560 11,214 3 12,331 19,188 4 17,881 25,585 5 25,618 31,959 6 31,921 No forecast 7 38,575 8 44,698 9 50,145 10 54,923 11 58,573 Initial Decision, 96 F.C.C.2d at 1198. Rogers’ demand figures represent its expected 40% market share. Rogers Direct Exh. 11, at 49; see Rogers Radiocall, Inc., 96 F.C.C.2d 1172, 1177 n. 13 (1984) (Chicago Final Decision, hereinafter Chicago). CMS’s fifth-year demand figures reflect its estimate of total market demand. CMS Direct Exh. II, at 14; see Chicago, 96 F.C.C.2d at 1177 n. 14. [10] Rogers also surveyed its current customers by mail, seeking requests for cellular units at a given cost for the proposed area; it received 1,284 responses to 9,238 inquiries. This survey apparently was not used in its demand calculations. ' [11] The actual need factor for Chicago could have been between 0.00375 and 0.00525, CMS Direct Exh. II, at Figure II — 7, which would yield an estimated range of demand from 26,633 to 37,286 units. [12] CMS was awarded a slight preference for its ability to accommodate demand for roamer service. CMS’s proposal was preferred because it was more convenient; its preference was only slight, however, because Rogers’ , superior grade of service provided greater excess capacity available to roamers. 96 F.C.C.2d at 1221. [13] Rogers Radiocall, Inc., 96 F.C.C.2d 1172 (Chi cago), reconsideration denied, 56 Rad.Reg.2d (P & F) 951 (1984). At CMS's request, oral argument in this case (and in the Pittsburgh proceeding as well) was heard before the Commission en banc on January 10, 1984. [14] The Commission added its reservations about the value of the comparative mode of decision-making in view of the cost to the applicants and the delay occasioned by the process. 96 F.C. C.2d at 1175. While there were sufficient differences between the proposals to choose comparatively a superior plan in view of the established criteria, the Commission observed that a basic, pre-comparative criterion was that all applicants be qualified to operate a cellular system. Id. at 1174. The Commission stated that the comparative selection of one did not brand the other as infeasible or incapable of providing cellular service to the public. Recognizing that either applicant could "reconfigure its system in order to meet the requirements of the marketplace" and that "the system designed to win the license may not be the system designed to serve the public," id. at 1174 & n. 5, the Commission nevertheless found itself "constrained to compare the applicants on the basis of the systems they have proposed, not on the basis of what might ultimately evolve.” Id. at 1174. These expressed doubts reflect not upon the ability of the Commission to select the superior application, given the established criterion, but upon the relative cost and delay incurred in utilizing a process which seeks to identify the "best” proposal. [15] The AU made separate findings and conclusions with respect to each sub-issue under Issue (a). The AU did not, however, award specific separate preferences for each sub-issue, instead incorporating all of his sub-issue conclusions in the single conclusion that Rogers deserved a substantial preference overall for Issue (a). On review, the Commission chose to award separate preferences for each sub-issue before determining the final Issue (a) preference. [16] The FCC reversed the preference awarded CMS for ability to accommodate demand for roamer service (service to cellular telephone subscribers from other systems). Because neither applicant demonstrated its superiority, no preference was awarded by the Commission. 96 F.C.C.2d at 1183-84. [17] In contrast, the successful applicant in Pittsburgh, MCI, received a slight preference under Issue (a) alone, which was held to outweigh the CMS affiliate’s slight preference under Issue (c). Pittsburgh, 96 F.C.C.2d at 1036. [18] The AU did not deny all opportunity to supplement the record. On at least three occasions, the AU requested that the parties or their expert witnesses provide further evidence such as detailed figures concerning geographic coverage, Order Prior to First Hearing Session (Nov. 5, 1982); geographic and population coverage data for CMS’s CGSA, Further Order Prior to First Hearing Session (Dec. 10, 1982); and testimony regarding CMS's equipment and five-year expansion plan, to be specified with particularity. Transcript at 46-49, 108-10, 137. [19] As an aside, CMS argues that the triviality of this distinction was further compounded by the admitted imprecision of the Carey contours; CMS, however, does not challenge the use of the Carey contours here. CMS Brief at 60-61; see Pittsburgh, at 187 n. 9 (Carey methodology and measurement of contours). [20] If CMS's coverage was indeed greater than that of Rogers, it was certainly afforded more than a sufficient opportunity to make that case before the Commission. In its Designation Order, the Commission found that CMS’s CGSA intruded into a central SMSA outside Chicago, a violation of the rules. CMS was permitted to amend its CGSA to conform. 91 F.C.C.2d at 518. Subsequently, the ALJ requested that the parties supply data for their proposed geographic areas and population coverage, which the ALJ was unable to find in the direct cases. Order Prior to First Hearing Session, at 2 (Nov. 5, 1982). Later, CMS was specifically asked to provide further data regarding its proposed coverage, as its prior effort was deemed "not responsive" to the earlier request. Further Order Prior to First Hearing Session (Dec. 10, 1982). Since CMS's CGSA extended four miles into Lake Michigan, which was not a part of the Chicago SMSA, CMS was also ordered to modify its CGSA to correct this error. Id. at 2. All parties were also asked to provide geographic and population data for their CGSA's as well. Id. at 2 n. 5. [21] The ALJ stated: CMS's contention that Rogers’ expanded system channel capacity and per-cell channel allocation are missing from the record are rejected. The number of voice channels in the Rogers system at the end of its first level cell split can be determined by reference to pages 156, 162 and 163-65 of Rogers Exhibit 1. Page 162 shows there will be 49 cells under Rogers’ first level split. Pages 163-165 show that 46 of these cells will be assigned three frequency groups (Cells 14 and 18 will require only two frequency groups each. Cell 16 will be assigned one group). Page 156 shows that each frequency group is assigned 15 voice channels, except groups 19, 20 and 21, which have 14 channels each. Thus, each cell is assigned 44 or 45 voice channels. Assuming, for ease of computation, cells are assigned 44 channels, the total number of channels assigned in Rogers’ system under the first level split is 2024 (46 cells x 44 channels), plus the channels assigned to cells 14, 16 and 18 (five 15 channel groups of 75 channels), for a total number of 2099 voice channels. Rogers Exhibit 1, Table IV-A reflects the frequency group assignments per cell during Rogers first level cell split and Table I-B reflects the chan *227 neis assigned to each frequency group. Rogers will assign specific channels from its channel allocation as needed. Initial Decision, 96 F.C.C.2d at 1208 n. 30.