Opinion for the Court filed by Chief Judge BAZELON.
BAZELON, Chief Judge:
At issue are Federal Communications Commission regulations dealing with one aspect of media ownership concentration: newspaper-broadcast station cross-ownership. Second Report and Order (Order), 50 F.C.C.2d 1046 (1974), reconsidered 53 F.C. C.2d 589 (1975).1 In 1970 the Commission proposed rules entirely eliminating newspaper-broadcast affiliation in a common city of operation. Docket 18110, 22 F.C.C.2d 349 (1970). Following a lengthy rulemaking proceeding, the Commission issued the more limited ban challenged here by the National Citizens Committee for Broadcasting, the Justice Department and various media interests.2 In relevant part, the rules:3 [7]*7Under the latter rule, roughly 90% of existing combinations are grandfathered. For the reasons expressed below, we affirm the prospective ban but vacate the rules dealing with existing combinations.
[6]*61) forbid the future formation or transfer of co-located newspaper-broadcast combinations4;
2) preserve existing combinations except those in communities served by only one daily newspaper and one broadcast station encompassing the community with a city grade signal; in the latter cases, one of the interests must be divested within five years.5
[7]*7I. THE RULEMAKING PROCEEDING AND ORDER
A. The Proceeding.
The Commission has long been aware that concentrated ownership threatens the public interest in a diversified mass communications media. The Commission’s first multiple ownership rules, adopted in the 1940’s, prohibited any party from owning or controlling more than one broadcast station in the same broadcast service in the same area.6 As a result of the AM radio rules, divestiture was ordered in 24 cases.7 The rules have since been tightened several times, most recently in the early stages of Docket 18110 to prohibit the common ownership of more than one broadcast station of any type in the same service area.8 The Commission enacted these rules without compiling a substantial record of tangible harm; it rested instead on its belief that “the effects of competition or its absence,” are “not readily susceptible of quantitative ascertainments.” 29 Fed.Reg. 7535, 7537, Docket 14711 (1967). The Commission has always viewed diversification as a factor of central importance in comparative license hearings.9
[8]*8In 1970 the Commission broadened the scope of ongoing rulemaking proceedings, Docket 18110, to consider whether newspaper-broadcasting affiliation was in the public interest.10 22 F.C.C.2d 339. Citing figures indicating the high incidence of newspaper-television affiliation,11 and the popular dependency on these media as information sources, the notice of proposed rule-making stated that, as with two television stations, it is not “desirable that these two organs ... be under the same control in any community.” Id. at 344, 346. The Commission suggested that promotion of media competition and maximum diversification of media viewpoints might require the separation of all co-located newspaper-broadcast combinations. Since the Commission contemplated allowing divestiture to take place over five years under favorable tax treatment,12 it expressed confidence that such a rule would produce no more than minimal economic losses. Id. at 347-48. Comments were invited.
Not surprisingly, the proposed rules generated considerable interest. Nearly 200 parties filed comments. In addition, roughly 25 major studies were lodged, Order at 1090-1094, although as will be explained in Section IV-A infra, the studies were of surprisingly little value.13 In March 1974, the Commission requested additional comments dealing primarily with the central problem of newspaper-television cross-ownership. 45 F.C.C.2d 768. Forty eight parties filed fresh comments.14 Three days of oral argument were held and the Order was released on January 31, 1975.
B. The Order.
After summarizing the comments and studies filed in Docket 18110, the Commission explained its justification for the prospective ban. Although the Commission had at one time encouraged newspaper sponsorship of broadcast stations, it concluded this policy was no longer warranted because broadcast licenses had become attractive to other types of investors. It had come to believe that “any new licensing [9]*9should be expected to add to local diversity,” Order at 1075. Because granting a broadcast license to a local newspaper owner could not be expected “to enhance diversity,” id., the Commission decided to prohibit the creation of future combinations. However, a geographic limit was placed on this ban because “there is no basis in fact or law for finding newspaper owners unqualified as a group for future broadcast ownership.” Id. Recognizing that various limits could be imposed, the Commission chose to employ the pattern used in earlier multiple ownership rules; the ban applies only if the station in question would or does encompass the common city with a city grade signal. Id,15
The Commission then focused on the more difficult question — what to do with existing co-located combinations. Initially, the Commission noted it had not yet paid sufficient consideration to certain possible side-effects of divestiture — reduction of local ownership, disruption of broadcasting continuity and local economic dislocations. Order at 1073. Claiming that stability and continuity of ownership serve the public interest, the Commission concluded that “a mere hoped for gain in diversity” was not adequate cause for disrupting ownership. Divestiture would be ordered “in only the most egregious cases.” Order at 1080. In framing its Order, the Commission chose not to focus, as the Justice Department suggested, on competition for advertising, citing authority that it has discretion to determine the weight to be given antitrust principles in shaping its regulatory policy.16 Instead, the Commission stated that the number of media voices serving a given area was a more important concern.
If our democratic society is to function, nothing can be more important than insuring that there is a free flow of information from as many divergent sources as possible. .
Order at 1079 (emphasis added).
The Commission then explained a combination would be considered an “egregious case” if it had an “effective monopoly in the marketplace of ideas” with respect to local issues. Although the Commission stated it had examined “every” combination “known to us,” Order at 1081, the record contained no evidence suggesting which combinations were monopolies in this sense. Instead, the Commission relied heavily on its experience, determining that a non-affiliated broadcast station encompassing the city of publication with a city grade signal could reasonably be expected to provide a sufficient additional voice on local matters, but that stations bathing the area with the weaker Grade A or Grade B signals could not.17 Thus, the Commission ordered divestiture only where no other newspaper or city grade broadcast station was operating in the locality of the combination.18
[10]*10As in its notice of proposed rule making, the Commission stressed that temporary or permanent waivers would be granted where appropriate because it did not intend to work a forfeiture through divestiture.19 It made clear, though, it would not reconsider in waiver proceedings issues resolved in Docket 18110. Order at 1085.20 Finally, the Commission indicated it did not intend to ignore antitrust considerations entirely, and would therefore continue to entertain petitions to deny on grounds of economic concentration. However, it tightened the requirements for granting such petitions, holding that it would set a hearing only when a violation of the Sherman Act was alleged. Order at 1088.21
The Commission reconsidered its regulations but did not materially alter them. 53 F.C.C.2d 589 (1975). The various petitioners then brought this appeal assailing the rules on numerous grounds.
The remainder of the opinion will be organized as follows, with the first page of each section being noted in parentheses.
II.Validity of the Prospective Ban (181 U.S.App.D.C. p. 10, 555 F.2d p. 947)
A. Whether the Prospective Ban Has a Rational Basis (181 U.S.App.D.C. p. 11, 555 F.2d p. 948)
B. Whether the Prospective Ban Exceeds the Commission’s Authority (181 U.S.App.D.C. p. 14, 555 F.2d p. 951)
C. Whether the Prospective Ban Violates the First Amendment (181 U.S. App.D.C. p. 16, 555 F.2d p. 953)
III. Whether the Commission May Adopt Any Rule That Orders Divestiture Without an Evidentiary Hearing (181 U.S.App.D.C. p. 18, 555 F.2d p. 955)
IV. Validity of the Divestiture Rule Adopted by the Commission (181 U.S. App.D.C. p. 19,555 F.2d p7 956)
A. Record Support for the Divestiture Order (181 U.S.App.D.C. p. 19, 555 F.2d p. 956)
B. Whether the Commission May Properly Require a Showing of Tangible Public Interest Harm to Support Divestiture (181 U.S.App.D.C. p. 24, 555 F.2d p. 961)
Conclusion (181 U.S.App.D.C. p. 29, 555 F.2d p. 966)
II. VALIDITY OF THE PROSPECTIVE BAN
The Commission must allocate use of the public airways by granting renewable licenses of no more than three years in the “public interest, convenience and necessity.” 47 U.S.C.A. §§ 301, 303, 309(a). Congress purposely left this mandate open-ended so that the Commission could fashion its regulations in light of changing conditions. Federal Communications Commission v. Pottsville Broadcasting Co., 309 U.S. 134, 138, 60 S.Ct. 437, 84 L.Ed. 656 (1940). In exercising its function, the Commission may enact eligibility standards for licenses. United States v. Storer Broadcasting, 351 U.S. 192, 76 S.Ct. 763, 100 L.Ed. 1081 (1955). Furthermore, the Commission may, under its authority to “encourage the larger and more effective use of radio in the public [11]*11interest,” 47 U.S.C. § 303(g), enact rules designed to curb potential abuses, Metropolitan Television Co. v. F.C.C., 110 U.S.App.D.C. 133, 289 F.2d 874 (1961).
Because the Commission has the primary responsibility for allocating use of the electromagnetic spectrum, a presumption of correctness attaches to its regulations. National Broadcasting Co. v. United States, 319 U.S. 190, 63 S.Ct. 997, 87 L.Ed. 1344 (1943). Of course, courts are not to rubber-stamp Commission regulations. On the contrary, review is to be “searching.” Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402, 415, 91 S.Ct. 814, 28 L.Ed.2d 136 (1971). This is especially so where the regulations touch on areas heavily laden with First Amendment interests.22 Yet despite this care, courts are not to substitute their judgment for that of the Commission or review the facts de novo. With respect to the prospective rules, which reflect a legislative type policy judgment,23 as opposed to the retroactive rules, which purport to be based also on factual considerations,24 our primary task will be to ensure that the regulations are not “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). In addition, we must examine the contentions that the rules exceed the Commission’s authority and violate the First Amendment.
A. Whether the Prospective Ban Has a Rational Basis
The licensing of speakers required by the Communications Act is unprecedented and has been justified under the First Amendment because of the scarcity caused b^ the physical limitations of the electromagnetic spectrum. The need for some regulation of the airwaves became clear in the 1920’s when there was none. “With everybody on the air, nobody could be heard.” National Broadcasting Co. v. F.C.C., 319 U.S. 190, 212, 63 S.Ct. 997, 1008, 87 L.Ed. 1344 (1943). In order to ensure the public’s ability to hear some speakers, the rights of other potential speakers were curtailed. The hard choice was between forcing free speech to bend or watching it break.
The licensing scheme was based on antitrust policies as well. It was feared that the conditions of scarcity would make broadcasting a natural monopoly so as to silence a diversity of opinion.25 “Congress moved under the spur of a widespread fear that in the absence of governmental control the public interest might be subordinated to monopolistic domination in the broadcasting field.” F.C.C. v. Pottsville Broadcasting Co., 309 U.S. 134, 137, 60 S.Ct. 437, 439, 84 L.Ed. 656 (1940).
The Commission has recognized that a policy of diversity is central to the Communications Act,26 and it has denied [12]*12licenses to otherwise qualified applicants on the strength of that policy.27 Since denying access to the airwaves to some involves a substantial though at this time necessary, restriction of speech, it follows that the Commission acts properly when it attempts to promote diversity by allocating stations to those without control over an alternative major media voice. The Supreme Court has given its approval to a diversity policy based on First Amendment and antitrust considerations. The First Amendment “rest on the assumption that the widest possible dissemination of information from diverse and antagonistic sources is essential to the welfare of the public.” Associated Press v. United States, 326 U.S. 1, 20, 65 S.Ct. 1416, 1424, 89 L.Ed. 2013 (1944).28 “The ‘public interest’ standard necessarily invites reference to First Amendment principles.” Columbia Broadcasting System, Inc. v. Democratic National Committee, 412 U.S. 94, 122, 93 S.Ct. 2080, 2096, 36 L.Ed.2d 772 (1972). In addition, even though the Commission does not enforce the antitrust laws, the Court has said the Commission may call on antitrust policies to promote diversity in giving content to the “public interest, convenience and necessity.” United States v. Radio Corporation of America, 358 U.S. 334, 351, 79 S.Ct. 457, 467, 3 L.Ed.2d 354 (1958).
Nevertheless there are limits to diversification policy.29 One limitation is contained in the statutory and First Amendment commands that the Commission shall not censor.30 The Supreme Court upheld the personal attack and political editorializing portions of the Commission’s fairness doctrine in Red Lion Broadcasting Co. v. F.C.C., 395 U.S. 367, 89 S.Ct. 1794, 23 L.Ed.2d 371 (1968). Those rules were designed to invigorate public debate by requiring broadcasters to air responses to its [13]*13programming in certain circumstances. The purpose of the rules is salutary.
It is the purpose of the First Amendment to preserve an uninhibited marketplace of ideas in which truth will ultimately prevail, rather than to countenance monopolization of that market.
395 U.S. at 390, 89 S.Ct. at 1806. Nevertheless, the fairness doctrine may well mark the outer limits of a permissible diversification policy which relies on direct government control over the content of broadcast programs.31 The Red Lion Court carefully withheld approval from rules that would more directly interfere with the broadcaster’s First Amendment rights, such as one restricting a broadcaster from airing its own views. 395 U.S. at 367, 89 S.Ct. 1794.32
In any event, the Constitutional difficulties with promoting diversity through speech restrictive means highlight the virtue of the prospective cross-ownership rules. The prospective ban is an attempt to promote diversity without government regulation of or supervision over speech. The rules attempt to promote vigorous public debate not by imposing restrictions on broadcasters, but simply by permitting more to be heard. Of course, there is no guarantee that the prospective ban will increase diversity in all cases, “the possible benefits of competition do not lend themselves to detailed forecast.” F.C.C. v. RCA Communications, Inc., 346 U.S. 86, 96, 73 S.Ct. 998, 1005, 97 L.Ed. 1470 (1952). But that does not make the attempt irrational.33 It is not unreasonable for the Commission to assume that the licensing of an independent, rather than an affiliated station, offers greater hope of providing the community a new voice and to incorporate that assessment in its licensing policy. The First Amendment seeks to further the [14]*14“search for truth.” Surely that search will be facilitated by government policy that encourages the maximum numbers of searchers.
B. Whether the Prospective Ban Exceeds the Commission’s Authority
The prospective rules, it is argued, cannot stand because, apart from their rationality, the Commission lacks authority to promulgate rules that restrict ownership of broadcast stations by newspapers. We hold that the prospective ban is within the Commission’s rulemaking authority.
Although no provision of the Communications Act expressly grants the Commission authority to restrict newspaper ownership of broadcast stations, such express authority is not essential to the validity of Commission regulations.34 47 U.S.C. § 309(a) requires the Commission to find the “public interest, convenience and necessity will be served” by the grant of a license. 47 U.S.C. § 303(r) grants the Commission authority to make such rules and regulations “not inconsistent with law, as may be necessary to carry out” the Act while 47 U.S.C. 154(i) grants rule-making authority to make rules “as may be necessary in the execution of [the Commission’s functions].”35 In United States v. Storer Broadcasting Co., 351 U.S. 192, 76 S.Ct. 763, 100 L.Ed. 1081 (1955), the Supreme Court upheld a portion of the Commission’s multipie ownership rules promulgated under these provisions.36 In essence, Storer permits the Commission to codify in rule its understanding, if reasonable, of the public interest licensing standard. (See 351 U.S. at 202-03, 76 S.Ct. 763). Here, the Commission has determined the public interest would not be served by granting broadcast licenses to newspapers and in Section II-A we held this policy was reasonable. It requires no extension of Storer to conclude the Commission possessed statutory authority to impose the prospective ban.37
Storer aside, ANPA contends the Commission is disabled from enacting this rule because the Commission’s first general counsel, Hampson Gary, was of the opinion that the Commission lacks the requisite statutory authority. In rejecting this contention, we need not determine what weight should be given to an agency’s initial interpretation of its mandate38 because Mr. Gary did not conclude that the Commission lacks statutory authority to enact a regulation restricting newspaper ownership. The question Mr. Gary addressed was whether “the Commission may deny an application of a newspaper for a license on the ground that it is against ‘public policy.’ ”39 After distinguishing “public policy” from the statutory standard of “public interest, convenience and necessity,40” he concluded that a license could not be denied on public [15]*15policy grounds but that newspaper ownership could bear on the public interest. Mr. Gary did not consider, as would be relevant here, whether the Commission may enact license qualifications in the public interest.41
A more substantial argument is that this court already has decided the Commission lacks authority to restrict newspaper ownership of broadcast stations. The case relied on most strongly is Stahlman v. F.C.C., 75 U.S.App.D.C. 176, 126 F.2d 124 (1942),42 where after holding that an F.C.C. subpoena was supported by the Commission’s general investigative powers, the Court wrote:
[The Commission’s power] does not embrace and should not be extended by implication to embrace a ban on newspapers as such . . . . This, we think, would be in total contravention of that equality of right and opportunity which Congress has meticulously written into the Act
126 F.2d at 127.
Although this language appears to resolve the question of statutory authority, in fact it does not. First, since the Commission had not then enacted a rule restricting broadcast ownership by newspapers, the issue of statutory authority was not before the court. Consequently, as Judge Edger-ton pointed out in his separate concurrence, the quoted language is dicta. 126 F.2d at 128.
More significantly, the premise of the paragraph containing the quoted language is that “[t]he Communications Act requires no more of an applicant for a radio license than proof of citizenship, character, and financial and technical qualifications to operate in the public interest.” Id. at 127. The next sentence implies that any applicant meeting these qualifications must be granted a frequency if one is available. However, this view of the Commission as a “traffic cop” of the airwaves was rejected by the Supreme Court the next year in National Broadcasting Co., supra. In order “to secure the maximum benefits of radio to all the people of the United States . Congress endowed the Communications Commission with comprehensive powers to promote and realize the vast potentialities of radio.” 319 U.S. at 217, 63 S.Ct. at 1010. Thus Stahlman was based on a reading of the Communications Act that has not prevailed.43
In sum, Storer supports the Commission’s exercise of rulemaking power and none of the materials cited requires a different conclusion.
[16]*16C. Whether the Prospective Ban Violates the First Amendment
It is argued that the prospective ban unconstitutionally conditions the First Amendment right to publish a newspaper.44 It is not argued, as indeed it could not be, that the ban violates the First Amendment rights of newspaper owners as would-be broadcasters. Denial of a license application does not abridge the applicant’s First Amendment rights. F.R.C. v. Nelson Bros. Bond & Mortgage Co., 289 U.S. 266, 53 S.Ct. 627, 77 L.Ed. 1166 (1933), National Broadcasting, supra, 319 U.S. at 226, 63 S.Ct. 997. “No one has a First Amendment right to a license or to monopolize a radio frequency; to deny a station license because ‘the public interest’ requires it ‘is not a denial of free speech.’ ” Red Lion supra, 395 U.S. at 389, 89 S.Ct. at 1806, quoting National Broadcasting, 319 U.S. at 227, 63 S.Ct. 997. Of course, it does not follow that the First Amendment has no role in the licensing system. The- Commission may not use the licensing process as a censoring mechanism, choosing those applicants whose views it favors.
Congress did not authorize the Commission to choose among applicants upon the basis of their political, economic or social views . . . If it did, or if the Commission by these Regulations proposed a choice among applicants on some such basis, the [First Amendment] issue before us would be wholly different.45
National Broadcasting, 319 U.S. at 226, 63 S.Ct. at 1014. But no claim of censorship can be made here. The Commission has only decided that the public interest is better served by assigning available broadcast frequencies to those without a co-located major media voice.
Freedom of the press does not necessarily shield newspaper publishers from regulation that may make publication more difficult. A newspaper could not, for example, plausibly claim the federal income tax violates the First Amendment because payment of the tax makes publication more difficult, or perhaps impossible. See Associated Press v. United States, supra. So the question presented is not whether the prospective ban burdens newspapers — it clearly may do so — but rather whether there is something special about the burden that renders it unconstitutional. The relevant case law does not support such a contention.
In Grosjean v. American Press Co., 297 U.S. 233, 56 S.Ct. 444, 80 L.Ed. 660 (1936), the Supreme Court held unconstitutional a state tax on net receipts imposed only on newspapers with weekly circulations of over 20,000. Justice Sutherland, writing for the Court, likened the tax to those imposed by the British on the American colonies in order to suppress dissent. Id. at 246-49, 56 S.Ct. 444. The tax was bad, he wrote because it was “a deliberate and calculated device ... to limit the circulation of information to which the public is entitled.” Id. at 250, 56 S.Ct. at 449. Here, the purpose of the Commission’s prospective regulation is precisely the opposite.46 The rule is intended and reasonably designed to increase the diversity of information and opinion circulated by increasing the number [17]*17of media voices in the community. Neither Grosjean nor the other cited cases are controlling.47
The ANPA admonishes us to heed the warning uttered by the Supreme Court in Miami Publishing Co. v. Tornillo, 418 U.S. 241, 256, 94 S.Ct. 2831, 41 L.Ed.2d 730 (1974);
Governmental restraint on publishing need not fall into familiar or traditional patterns to be subject to constitutional limitation ... Id. at 256, 94 S.Ct. at 2839.
However, a restraint is not unconstitutional merely because it is unfamiliar. Unable to claim the ban violates their First Amendment rights as would-be broadcasters, petitioners have not demonstrated the ban trenches on the constitutionally protected interests of newspaper publishers.48 The ban neither mandates nor prohibits what may be published. It only forbids the formation of co-located combinations the Commission has determined do not serve the public interest. Of course, in serving the public interest the ban may indirectly affect the economics of publication, but the First Amendment does not provide a shield against this. Associated Press, 326 U.S. at 20, 65 S.Ct. 1416.
In fact, the very claim of constitutional abridgement is ironic. First of all, the prospective ban is an attempt to enhance the diversity of information heard by the public without on-going government surveillance of the content of speech. Second, it may be that newspapers can not truly be free of government interference so long as they operate government licensed broadcast stations. An unsavory fact of life is that government has the power to regulate expression by a “raised eyebrow” reminding the broadcaster of the triennial government renewal process. A newspaper opens itself up to similar intimidation by affiliation with a broadcast station.49 “It would be strange indeed, however, if the grave concern for freedom of the press which prompted adoption of the First Amendment should be read as a command that the government was without power to protect that freedom. . . . ” Associated Press, 326 U.S. at 20, 65 S.Ct. at 1424.
* * % * if: *
In conclusion, we affirm the validity of the prospective ban. We find that the ban is a rational attempt to promote the highly valued goal of diversity, that it is within the Commission’s authority and that it does not violate the Constitution. We now turn to the issues regarding the Commission’s treatment of existing combinations. How[18]*18ever, before considering whether the divestiture scheme actually ordered by the Commission can be upheld, we must evaluate the arguments that the Commission may not enact any rule that orders divestiture without an evidentiary hearing.
III. WHETHER THE COMMISSION MAY ADOPT ANY RULE THAT ORDERS DIVESTITURE WITHOUT AN EVIDENTIARY HEARING
From the premise that divestiture, as retroactive legislation, is inherently suspect, it is argued that divestiture is permissible only if Congress specifically intended for the Commission to enact retroactive rules and only where, quoting from the Order at 1083, “the need is overwhelming and the evidence is unambiguous.”50 This argument fails in light of American Airlines, Inc. v. Civil Aeronautics Board, 123 U.S.App.D.C. 310, 359 F.2d 624 (1966), where in upholding CAB regulations that affected existing licenses, this court en banc rejected an argument “that the Storer doctrine is restricted to regulations affecting applications for new licenses or certificates . ” Id. at 628. The court said “such rule making is not to be shackled, in the absence of clear and specific Congressional requirement.” Id. at 629. Authority to order divestiture follows not only from the fact that American Airlines is based on Storer Broadcasting, which interpreted the Communications Act, but also from the fact that the interest in a broadcast license is limited and conditional. A licensee must apply for'renewal of authority every three years, and the renewal application is subject to such rules and conditions as are in effect, not during the license period, but at the time of the renewal. Transcontinent Television Corp. v. F.C.C., 113 U.S.App.D.C. 384, 308 F.2d 339 (1962). Consequently, the contention that divestiture contains elements common to suspect retroactive legislation goes not to the existence of Commission authority, but rather to the reasonableness of its exercise.51
Section 309(e) of the Act, 47 U.S.C. § 309(e), requires the Commission to grant an applicant a hearing if unable to make the public interest finding required to grant a license. NAB advances the argument, seemingly foreclosed by Storer and National Broadcasting, that a divestiture rule violates the statutory right to a hearing. Although it is not entirely clear, the basis of this argument appears to be that Storer and National Broadcasting apply only to potential licensees not existing ones. However, as American Airlines suggests, these cases are not dependent on a distinction between current and potential licensees and nothing presented to us suggests why this distinction should be read into them.52 Further[19]*19more, divestiture would not be as harsh as denial of license renewal. A licensee facing divestiture is compensated for its loss. A licensee that fails of renewal is not.
While recognizing that the Commission has broad discretion to proceed by rulemaking or adjudication, and that it may affect the rights of individual licensees without first holding an evidentiary hearing, Gray Broadcasting argues that the Commission abused its discretion by ordering divestiture without affording affected parties a prior evidentiary hearing. Gray contends the Commission has couched as rulemaking a proceeding that in “substance and effect is individual in impact,” in other words an adjudication. The APA defines rule as “any agency statement of general or particular applicability and future effect designed to implement, interpret, or prescribe law or policy.” 5 U.S.C. § 501(c). Professor Davis points out that the words “or particular” were inserted after Senate passage “to assure coverage of rule making to named persons.” Davis, 7 Administrative Law § 5.02 (1958). The nature of the evidence required to implement a rule, and not the particularity of application, determines the necessity for individualized hearings. Given our disposition of the case, however, we need not determine whether a divestiture order limited to the seven affected licensees could have been implemented without individualized hearings.53
In sum, the Commission has authority to promulgate a rule that orders divestiture. An altogether different question, to which we now turn, is the reasonableness of the scheme under review.
IV. VALIDITY OF THE DIVESTITURE RULE ADOPTED BY THE COMMISSION
Since the Commission purported to limit divestiture to “egregious cases” of “effective monopoly,” we must examine the record support for this determination. Greater Boston Television Corp. v. F.C.C., 143 U.S.App.D.C. 383, 444 F.2d 841, 850 (1970). The purpose of our review is only to determine whether the agency decision is rational and based on consideration of the relevant factors, Citizens to Preserve Over-ton Park v. Volpe, supra, 401 U.S. at 146, 91 S.Ct. 814; Bowman Transportation v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 285, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974); whether, in terms of the APA, the agency decision is arbitrary and capricious. 5 U.S.C. § 706(2)(A). Technically, there is no requirement that the rules be based on substantial evidence. Cf. 5 U.S.C. § 706(2)(E). Nevertheless, a decision must be considered “arbitrary and capricious” if the facts on which it is purportedly based are not supported by the record. Associated Industries of New York State, Inc. v. U. S. Dept. of Labor, 487 F.2d 342, 349-50 (2d Cir. 1973) (Friendly, J.). Factual certainty is not necessary, though. An agency may regulate even though the facts do not illuminate a clear path. Ethyl Corporation v. EPA, 176 U.S.App.D.C. 373, 541 F.2d 1 (1976). But it stands to reason that presumptions derived from the enabling statute and fundamental national policies take on redoubled importance where this is so.
A. Record Support for the Divestiture Order
The voluminous record of the lengthy rulemaking proceeding contains little reliable “hard” information. Many filings are devoted to legal argument or representations that the editorial staff of the filer’s newspaper and television station are inter[20]*20nally segregated. In addition, roughly twenty-five major studies were submitted.54 However, as the Commission brief points out, the studies were largely inconclusive. Br. at p. 9.
We summarize the record only to the extent necessary to show that in determining its divestiture policy, the Commission necessarily had to rely primarily on policy, not factual considerations.55
(1) Effect of Common Ownership on Station Performance
Several studies attempt to demonstrate either that newspaper ownership of television stations has not harmed the public interest or, affirmatively, that newspapers have been superior broadcast licensees. The Anderson study56 is largely of the former variety. It attempts to compare patterns of content under different ownership conditions to determine if cross-ownership produces “repressive information control,” evidently on the assumption that the purpose of diversity policy is to combat such repression. The study finds no significant difference in content between individually and cross-owned stations in the sample cities and concludes divestiture is unnecessary.57 Although the study has an objective format, it is based on content analysis that necessarily entails subjective judgments.58
The Litwin and Wroth study,59 commissioned by the NAB, compares media performances in highly concentrated and relatively unconcentrated cities. Relying on information acquired in interviews, it finds that cross-owned stations are perceived to provide greater amounts of news coverage in greater depth than individually owned stations'. In part, it attributes this superiority to greater financial stability. However, the interviewers only polled high-ranking “media personnel” and “business and community leaders.”60 Furthermore, the difficulties with an approach based on popular perceptions is highlighted by the fact that the study reports a pervasive history of news distortion and bias in one of the three highly concentrated cities examined.61
The Commission undertook a study62 that compares the performance of cross and individually owned stations, but without making judgments based on content analysis. A complex multiple regression analysis finds that cross-owned stations present “6% more local news, 9% more local non-entertainment, and 12% more total local including entertainment than do other TV stations.” This study, the Order concludes, shows “an undramatic but nonetheless statistically significant superiority in newspaper owned television station in a number of program particulars.” Order at 1078 n. 26. However, the Commission did not stress this superiority in deciding to grandfather 'most stations. Comparing amounts of programming types aired, not only requires qualitative judgments in categorizing the programs, but would also seem to be a [21]*21crude measure of the public interest. More broadly, as Commissioner Robinson asserted in his dissent, it is inconsistent for newspapers to assert that they make superior broadcasters while at the same time claiming that their affiliated stations are totally independent. Order at 1119.
An empirical study which employed data drawn from the Commission’s files attempts to determine if cross-owned stations differ from other stations in the amounts of news and other public interest programming they air.63 According to the Commission, this study finds “no significant differences.” A study by Harvey Levin64 reaches the opposite conclusion. He finds that individually owned stations “systematically outclass” affiliated stations in their levels of public interest programming. However, several filings severely criticize Levin’s methodology65 and, although he later attempted to correct the deficiencies,66 the Commission termed his results “inconclusive.” Order at 1073.
Finally, the American Institute conducted a study to determine whether cross-ownership affects the level of public knowledge.67 It finds the public in Zanesville, Ohio, a town with media “monopoly,” significantly less informed than the public of a “similar” city, York, Pennsylvania, a town with a competitive media. The Institute attributes this difference to media monopoly. It is possible, however, that factors other than media monopoly contribute to or explain the results. In fact, this study highlights the difficulty of attempting to compare the performances of cross-and independently-owned stations. As Commissioner Robinson observed in his dissent:
We ought not to assume that the kind of “goodness” we really want in a station is measured simply by diversity among different program types, the relative percentage of situation comedies, news, public affairs and the like. On the other hand, it is about the only way we can get any kind of objective handle on program quality, without getting into the business of making subjective valuations of programming content, a course that, needless to say, raises grave First Amendment problems.
Order at 1121.
(2) Effect of Common Ownership on Competition
Comparing advertising rates provides the most ready means for determining whether cross-ownership in general has an anti-competitive effect. Rosse et al.68 purport to find that newspaper-owned television stations charge advertising rates 15% higher than other stations, presumably due to their monopoly position. However, their calculations assume that audience size lacks explanatory value because it is a “dependent variable”; in other words, they assume that since advertising prices determine programming quality which in turn determines audience size, audience size cannot explain advertising rates.
In a NAB-commissioned study, Lago and Osborne challenge this assumption.69 Employing audience size as an explanatory variable in a regression analysis, this study finds no significant difference in advertis[22]*22ing rates. The Commission did not indicate which methodology it accepted; it merely noted that its own examinations failed to confirm Rosse’s conclusion.70 Order at 1073.
Apart from these studies, the Commission noted that the record did not contain evidence of specific anti-competitive acts by cross-owned stations. However, Stephen Barnett suggested before the Commission that such evidence — as well as evidence regarding the impact of cross-ownership on station performance — is available, though difficult to uncover.71 Pointing to examples from the Commission’s cases, the Lit-win and Wroth report 72 and other documented sources, Barnett described several instances where multiple-media owners had abused their power. Typical abuses were news suppression, news distortion and advertising discrimination that presumably would have been less feasible in a more competitive market.
The Commission did not mention these or other examples of abuses in its summary of the record73 although Commissioner Hooks noted that Barnett’s position was supported by several other submissions. Order at 1109 (Hooks, Comm., concurring in part). The Commission’s focus was on national trends, and it may have felt uncomfortable drawing conclusions about an entire industry from apparently isolated examples. However, former Commissioner Johnson has expressed the opinion that the majority of media abuses remain undiscovered. In re National Broadcasting Co., 19 F.C.C.2d 713, 720-723 (dissenting opinion). In his view, this is caused by lack of written records, the difficulty of separating the elements of self-interest and news judgment, and the fact that station managers rarely resort to overt, documentable pressures.
(3) Multiplicity and Diversity of the Media
The record contains several studies designed to measure media diversity. Sterling’s study reports that the proportion of newspaper-owned broadcast stations has declined since 1950.74 However, it examines only the top 100 markets. Furthermore, it may understate the amount of current concentration. In gauging concentration, radio and television stations are equated even though the Commission had noted in its notice of proposed rulemaking that TV and newspapers are by far the most powerful media sources.75 Thus the study would show that a town originally served only by a cross-owned television-newspaper combination had become significantly less concentrated by the addition of a radio station, even if the co-located combination remained the dominant voice. Udell’s study reaches a conclusion similar to Sterling’s. It finds that the percentage of newspaper-owned television stations dropped from 29% in 1955 to 14% in 1969.76 This study may simply reflect, however, the fact that the number of broadcast outlets increased during this period while the number of newspapers decreased, a finding also made by Sterling.77
[23]*23The Seiden study attempts to measure precisely the amount of diversity in communities throughout the country.78 Choosing the Area of Dominant Influence79 as the relevant geographic area instead of either the city, country or Standard Metropolitan Statistical Area, it counts the number of media competitors and owners that serve' each area and finds that no area is served by less than thirty media owners.80 However, the count includes numerous media that have little or no bearing on local diversity, such as newspapers from other cities and national magazines. The count contains other distortions as well. A television station and its satellites are treated as individual outlets even though they provide the same service. The Commission characterized this study as “totally unrealistic.” Order at 1079, n.28.
In concluding its discussion of these studies, the Commission expressed doubt as to whether “counts” as such generated valuable knowledge. Order at 1074. Also, the Commission’s brief disparages the conclusion of the Justice Department brief that the media is highly concentrated in many cities. In thirty Standard Metropolitan Statistic Areas (SMSAs) involved in this case, Justice observes, only one newspaper circulates. “In twenty others there are only two daily newspapers, and in eleven of these the newspaper which owns a television station has more than 75% of the SMSA’s newspaper circulation. Computed from Rand Corporation Study R-1585MF.” And, Justice continues, “metropolitan areas which in theory have several newspapers, in practice usually have only one or two with significant circulation.” Br. at 20. Such figures, the Commission points out, do not establish that the broadcast industry is anti-competitive.
(4) Economic Consequences of Divestiture
The economic consequences of divestiture involves two questions: the extent to which investors themselves will be injured and the extent to which this injury will be translated into public injury. Two major studies deal chiefly with the first question, none with the second.
Commissioned by the ANPA, the FrazerGross study81 and the First National City Bank study82 both suggest most combination owners will divest their television stations, at less than current value. In particular, Frazer estimates that under divestiture conditions, TV stations will sell at 10% to 20% less than true value. In response, Justice contends this conclusion is premised on the erroneous assumption that broadcast licensees do not compete in the investment market with other non-broadcast income-producing properties.83 Furthermore, even assuming the validity of Frazer’s judgment, it does not follow that the sales price will not include a substantial profit margin.84 [24]*24Finally, since investment demand and license demand may fluctuate during the five year sale period, it is difficult to forecast loss figures accurately. The Commission concluded, “Our own evaluation is that the concern [over losses] may be exaggerated.” Order at 1072.
No studies dealt with whether the public would be injured by the economic consequences of divestiture. Various comments, however, predict divestiture will produce harmful side-effects such as reductions in local ownership or the demise of marginally profitable newspapers contrary to the intent of the Newspaper Preservation Act.85 Such suggestions are little more than speculations.86
B. Whether the Commission May Properly Require a Showing of Tangible Harm to Support Divestiture
Putting aside for the moment the question whether the Commission abused its discretion in refusing to order divestiture absent a showing of tangible public interest harm, we first consider whether the Commission’s evaluation of the record was reasonable.
Despite the fact that it did not systematically evaluate the record, the Commission evidently concluded that the factual case for broad-scale divestiture had not been made. At one point the Order implies that the Justice Department failed to demonstrate that cross-owned stations charged anti-competitive advertising rates, Order at 1079 n. 27, and later it observes “it might have been necessary ... to require divestiture in many more situations if the power of the print-broadcast combination were exercised monolithically.” Order at 1089. Also, after mentioning its power to act in the absence of documented abuses, the Commission nonetheless concluded that such “considerations are [not] a substitute for the requirement of a stronger showing than we have of the need for such a severe remedy on a broad basis.” Order at 1080, n.29.
Furthermore, the Commission reasonably could have concluded, as it evidently did, that a compelling factual showing was unlikely to be made. The Commission not only received roughly 200 original submissions, but, when dissatisfied with the substantially of these filings, there were 48 fresh filings in response to a detailed request. 45 F.C.C.2d 768. As a whole, the submissions suffered from what may have been unavoidable failings.87 Diversity and its effects are, as Commissioner Robinson suggested, elusive concepts, not easily defined let alone measured without making qualitative judgments objectionable on both policy and First Amendment grounds. And the process suggested by Professor Barnett’s filing did not offer much more hope of compiling a useful record. Former Commissioner Johnson has suggested that specific abuses of cross-ownership are unlikely to be apparent. And, as Judge Tamm has explained elsewhere, documentation of such abuses, “would require not only program monitoring but also laborious perusal of reams of documentary evidence which will likely be dispersed, inaccessible and unintelligible to one who is not skilled in the intricacies of antitrust and communications law.” Hale v. FCC, 138 U.S.App.D.C. 125, 425 F.2d 556, 565 (1970) (Tamm, J., concurring). Furthermore, the intrusiveness involved in the very process of attempting to uncover abuses — such as news distortion— might severely strain First Amendment values.
On the other hand, the Order does not emphasize that the record does not support the reasons adduced for grandfathering either. There is virtually no evidence regarding the likelihood that divestiture would [25]*25produce the harmful effects feared, or how serious any such harm would be. And the record no more establishes that cross-ownership serves the public interest than injures it.88 After years of study, the record was essentially inconclusive.89
Had the evidence pointed strongly one way or another, the Commission probably would not have considered whether it should order divestiture absent a showing of tangible harm. Having concluded that the record was inconclusive, the Commission recognized that it nonetheless had the power to order divestiture. Order at 1080, n.29. Had the Commission reasoned that because diversified ownership is prima facie in the public interest, local cross-ownership violates the public interest, it would have been rational for it to conclude that divestiture was generally called for, subject to any appropriate exceptions. The Commission did not adopt this view, however,90 “Divestiture,” the Commission wrote, “is a harsh remedy, one to be reserved [sic] only where the need is overwhelming and the evidence unambiguous.”91 Id. at 1083. Thus the question before us is whether the Commission acted properly in refusing to order divestiture in the absence of a showing of tangible public interest harm.
Having evaluated the policies and record presented to us, we hold that the Commission erred in concluding that the other policies it advanced required it to find evidence of harm before ordering divestiture. Divestiture, to begin with, is a misleading term in this context. It implies that the broadcaster has that which the Communications Act specifically states he does not have — an interest in the license beyond its expiration date — and that he is being forcibly deprived of a vested right. A licensee must apply for renewal of his license every three years and the Commission is to grant renewal only if it finds it is in the public interest to do so. 47 U.S.C. § 307(d).
With this in mind, it is easier to understand that even though divestiture is typically a remedy for misconduct, it is actually a much more flexible tool. Divestiture need not only be used to take the licenses from those who disserve the public interest. As with the Commission’s early Chain Broadcasting and Multiple Ownership rules,92 divestiture could be employed to increase the extent to which the broadcasting industry serves the public interest. As the Commission recognized, “it is unrealistic to expect true diversity from a commonly owned station-newspaper combination. The divergency of their viewpoints cannot be expected to be the same as if they were antagonistically run.” Order at 1079-80. Although divestiture cannot guarantee greater diversity, it increases the likelihood that the public will be served by broadcasters with diverse views. Thus the Commission’s obligation to “encourage the larger [26]*26and more effective use of radio in the public interest,” 47 U.S.C. § 303(g) would support the presumption that cross-owned stations do not serve the public interest.
The policies of the First Amendment also support this presumption. Because the right of access to the airwaves of some would-be broadcasters has been restricted, the public is deprived of speakers it would otherwise hear. As a result, the Supreme Court has said “[i]t is the right of the viewers and listeners, not the right of the broadcasters, which is paramount.” Red Lion, supra, 395 U.S. at 390, 89 S.Ct at 1806. And the interest of the public is presumed to be in maximum diversification of the airwaves. The First Amendment “presupposes that right conclusions are more likely to be gathered out of a multitude of tongues, than through any kind of authoritative selection. To many this is, and always will be folly; but we have staked upon it our all.” Associated Press v. United States, 52 F.Supp. 362, 372 (D.C. 1943).
Finally, and perhaps most importantly, the Commission’s long-standing policy has been that diversification of media sources is of central importance.93 It reaffirmed this view throughout the proceeding under review. “If our democratic society is to function, nothing can be more important than insuring that there is a free flow of information from as many divergent sources as possible.” Order at 1079 (emphasis added). Of course, the Commission probably did not intend for this and similar statements to be read literally. If it had, it presumably would not have allowed itself even to consider whether other policies ever called for waiver. But at the least, consistency would call for ordering divestiture absent a showing of greater public interest harm.
The Commission did not do this, however. Although purporting to reaffirm the primacy of diversity, e. g., Order at 1078, the Order ultimately did not adopt the presumption that cross-ownership does not serve the public interest, divestiture only being ordered where “the need is overwhelming and the evidence unambiguous.” Order at 1083. The Commission never explained that or why it had abandoned the presumption suggested in the notice of proposed rulemaking. It simply cited the existence of competing policies it claimed not to have previously considered.
The first competing value is preservation of local ownership and involvement of ownership in management. The Commission expressed fear that “many sales would have to be to outside interests.” Order at 1078. However, as the Policy Statement on Comparative Broadcast Hearings, 1 F.C.C.2d 393, 395-96 (1965), makes clear, local ownership by itself is of little concern to the Commission 94 Owner management is significant because it increases the likelihood that station policy will be attuned to local needs; local ownership by itself offers no such promise.95 And even when both are present, they are not of primary importance to the Commission. A licensee may sell its license without regard to whether the new owner intends to participate actively in management. Br. of Justice Dept, at 39, F.C.C. Reply Br. at 19 n.15.
In addition, the record does not support the theory that divestiture will impair local ownership and management. The Justice Department asserts and the Commission does not deny that nineteen, or roughly one quarter of the 79 newspaper-television com[27]*27binations at issue, are not locally owned. In those cases divestiture may spark improvement. Furthermore, no evidence was presented that the local owners of the other stations are actively involved in daily management. Since “[m]ost of the parties state that their broadcast stations and newspapers have separate management, facilities, and staff . . . ” Order at 1059, it may be that owner management is unusual. Finally, there is no reason to suppose that local entrepreneurs will not find television an attractive investment.96
The second fear is that divestiture might disturb “continuity of operation as the new owner would lack the long knowledge of the community and would have to begin raw.” Order at 1078. As support, the Commission brief cites judicial decisions giving recognition to the fact that “legitimate renewal expectancies are implied in the structure of the Communications Act” Rep.Br. at 15. The goal of certainty is to give broadcasters an incentive to make the investment required to produce quality programming. Whatever the legitimate strength of this policy in light of the statutory limitation on the interest in a license,97 it has scant force in this case. A one-time alteration in the ownership structure of the broadcast industry should not effect the public’s interest in quality programming because the new owners will become the beneficiaries of any unofficial policy of continuity.98 Furthermore, to the extent the Commission means that transfer of ownership is by itself disruptive, its position is at odds with its practice of routinely approving license assignments.99 Finally, even if transfer to non-local owners entails extraordinary disruption, there is no basis for assuming that such transfers will result from divestiture.100
The final fear is that “[l]ocal economic dislocations are also possible as a result of the vast demand for equity capital and wide-scale divestiture could increase interest rates and affect selling price too.” Order at 1083. Although it is not entirely clear, the Commission presumably was not concerned with financial hardship to the individual owners.101 Private losses are a relevant concern under the Communications Act only when shown to have an adverse effect on the provision of broadcasting service to the public. Federal Communications Commission v. Sanders Radio Station, 309 [28]*28U.S. 470, 474-76, 60 S.Ct. 603, 84 L.Ed. 869 (1940); Carroll Broadcasting v. F.C.C.., 103 U.S.App.D.C. 346, 258 F.2d 440 (1958). Yet, as Commissioner Robinson theorized in his dissent, Order at 1127, divestiture should not even produce substantial private losses. Most cross-owned stations are held by their original owners who by now have long recouped their original investment. Order at 1130.102
In its brief, the Commission claims that its concern was the effect divestiture would have on future programming. It reasoned that because broadcasting stations are expensive, new owners would have insufficient working capital to finance news and other public affairs programming. Rep.Br. at 20. But presumably this would also be the case in voluntary transactions, which the Commission readily permits. A different public interest injury that might result from separation is the failing of marginal newspapers no longer supported by affiliated broadcast stations.103 However, the Commission did not rely on this point and, as Commissioner Robinson pointed out, the record contains no reliable information on the degree to which stations subsidize newspapers. Order at 1129.104
In sum, the Commission could not have rationally concluded that the competing policies it offered justified grandfathering absent a show of harm. And, since the record does not disclose the extent to which divestiture would actually threaten these values, the Commission could not have rationally concluded, if it did, that their potential impairment overcame the presumption against cross-ownership.105 The gains in diversity from divestiture may be speculative, but since divestiture is the most promising method for increasing diversity that does not entail governmental supervision of speech,106 the Commission could not rationally conclude that lesser policies, lacking , support in the record, require maintenance of the status quo.107 “In passing upon considerations open to the Commission even on an application for renewal of a license this [29]*29court has taken a stand against heavy reliance upon maintenance of the status quo.” TV 9, Inc. v. F.C.C, 161 U.S.App.D.C. 349, 495 F.2d 929, 937 (1973).108
In any event, even if the Commission’s chosen presumption were supportable, its current policy could not stand. First, the record does not support the conclusion that divestiture would be more harmful in the grandfathered markets than in the 16 affected markets. Nor does it show that the need for divestiture in the affected markets is “overwhelming” and the evidence “unambiguous.” Although the affected markets contain fewer voices, the amount of diversity in communities with additional independent voices may in fact be no greater. The Commission “must do more than enumerate factual differences, ... it must explain the relevance of those differences to the purposes of the Federal Communication Act.” Melody Music v. F.C.C., 120 U.S.App.D.C. 241, 345 F.2d 730, 733 (1965).
Finally, limiting divestiture to small markets of “absolute monopoly” squanders the opportunity where divestiture might do the most good. There is no apparent technological reason why additional broadcasters could not be licensed in many of the smaller markets.109 In many large markets, on the other hand, newspapers possess highly coveted VHF licenses.110 Divestiture thus may be more useful in the larger markets. The Commission’s arbitrary cut-off at one independent in-coming signal prevents realization of such gains.111
Conclusion
Although we do not disturb the Commission’s prospective rules, we conclude that the divestiture order is inconsistent with its longstanding policy that “nothing can be more important than insuring that there is a free flow of information from as many divergent sources as possible.” The Commission has sought to limit divestiture to cases where the evidence discloses that cross-ownership clearly harms the public interest. For the reasons expressed above, we believe precisely the opposite presumption is compelled, and that divestiture is required except in those cases where the evidence clearly discloses that cross-ownership is in the public interest.112 Accordingly, we vacate such portions of the Order that have retroactive effect and remand the [30]*30record to the Commission for adoption of a rule not inconsistent with this opinion.
It is so ordered.