Opinion for the Court filed by Circuit Judge STEPHEN F. WILLIAMS.
Opinion concurring in part and dissenting in part filed by Chief Judge WALD. Concurring opinion filed by Circuit Judge STARR. STEPHEN F. WILLIAMS, Circuit Judge:Under the “fairness doctrine,” the Federal Communications Commission has, as its 1985 Fairness Report explains, required broadcast media licensees (1) “to provide coverage of vitally important controversial issues of interest in the community served by the licensees” and (2) “to provide a reasonable opportunity for the presentation of contrasting viewpoints on such issues.” Report Concerning General Fairness Doctrine Obligations of Broadcast Licensees, 102 F.C.C.2d 143, 146 (1985). In adjudication of a complaint against Meredith Corporation, licensee of station WTVH in Syracuse, New York, the Commission concluded that the doctrine did not serve the public interest and was unconstitutional. *40Accordingly it refused to enforce the doctrine against Meredith. Although the Commission somewhat entangled its public interest and constitutional findings, we find that the Commission’s public interest determination was an independent basis for its decision and was supported by the record. We uphold that determination without reaching the constitutional issue.
I.
In the summer of 1982 Meredith ran a series of advertisements over WTVH arguing that the Nine Mile II nuclear power plant was a “sound investment for New York.” Syracuse Peace Council complained to the Commission that Meredith had failed to give viewers conflicting perspectives on the plant and had thereby violated the second of the fairness doctrine’s two requirements.
In its initial decision the Commission agreed with Syracuse that Meredith had failed to fulfill its obligations under the doctrine and demanded that the station within 20 days give notice of how it planned to meet those obligations. Syracuse Peace Council, 99 F.C.C.2d 1389, 1401 (1984).
Meredith filed a petition for reconsideration in which it argued that the fairness doctrine was unconstitutional. Meredith Reply to Opposition and Petition for Reconsideration and Supplement, April 12, 1985, at 11-41, Joint Appendix in Case No. 85-1723 at 261-91. Before ruling on Meredith’s petition, the Commission completed its 1985 Fairness Report, the culmination of a separate inquiry into both the wisdom and constitutionality of the fairness doctrine. Inquiry into Alternatives to the General Fairness Obligations of Broadcast Licensees, 102 F.C.C.2d 143 (1985).
On the issue of whether the doctrine continued to promote the public interest, the 1985 Report said that the Commission was “firmly convinced that the fairness doctrine, as a matter of policy, disserves the public interest ...” 102 F.C.C.2d at 148. See also id. at 147 (similar); id. at 246 (language identical to that quoted). Ir. reaching that conclusion the Commission invoked essentially the same grounds as it has in the present action — chiefly, that growth in the number of broadcast outlets reduced any need for the doctrine, that the doctrine often worked to dissuade broadcasters from presenting any treatment of controversial viewpoints, that it put the government in the doubtful position of evaluating program content, and that it created an opportunity for incumbents to abuse it for partisan purposes. Despite all this, it declined to eliminate the doctrine, expressing concern that it might be statutorily mandated. Id. at 148.
The 1985 Report also raised serious doubts about the continuing constitutionality of the fairness doctrine, but, saying that it was “the province of the federal judiciary —and not this Commission — to interpret the Constitution,” 102 F.C.C.2d at 155, the Commission refused to make a constitutional ruling.
After issuing the 1985 Report, the FCC in due course considered Meredith’s petition for reconsideration. In that context it again refused to address the constitutional issue — not on the ground that Meredith had raised the defense belatedly but solely on the theory that that issue should be left to Congress and the courts. It invoked its 1985 Report in support of this view. Syracuse Peace Council, 59 Rad.Reg.2d 179, 182 n. 4.
On appeal, this court reversed and remanded the case to the Commission. Meredith Corp. v. FCC, 809 F.2d 863 (D.C.Cir.1987). We noted the principle that regulatory agencies cannot invalidate an act of Congress, see Johnson v. Robison, 415 U.S. 361, 368, 94 S.Ct. 1160, 1166, 39 L.Ed.2d 389 (1974), but said that an agency could not blind itself to a constitutional defense to a “self-generated” policy. 809 F.2d at 872. In the meantime, we observed, this court had in another decision found that the fairness doctrine was not mandated by statute. Telecommunications Research & Action Center v. FCC, 801 F.2d 501, reh ’g en banc denied, 806 F.2d 1115 (D.C.Cir.1986) (“TRAC”), cert. denied, 482 U.S. 919, 107 S.Ct. 3196, 96 L.Ed.2d 684 (1987). See 809 F.2d at 873 n. 11. Thus on remand *41the Commission would have to resolve Meredith’s constitutional defense, unless it determined in light of the 1985 Report that enforcement of the doctrine was contrary to the public interest. Id. at 874. We explicitly noted that the 1985 Report had concluded that indeed the doctrine no longer served the public interest standard of the Communications Act, see id. at 867, and that in the Report the Commission had “largely undermined the legitimacy of its own rule” and “eviscerate[d] the rationale” for the doctrine, id. at 873.
On remand, the Commission expanded the scope of the Meredith proceeding by soliciting comments from the public on the general questions whether “in light of the 1985 Fairness Report, enforcement of the fairness doctrine is constitutional and whether enforcement of the doctrine is contrary to the public interest.” Syracuse Peace Council, 52 Fed.Reg. 2805, 2805 (Jan. 27, 1987). In its Memorandum Order and Opinion in Syracuse Peace Council, 2 F.C.C.Rcd. 5043 (1987), recon. denied, 3 F.C.C.2d 2035 (1988), the Commission ruled in favor of Meredith.
The FCC relied heavily on the conclusions drawn in the 1985 Fairness Report, and in fact incorporated that Report into the record and “reaffirm[ed] [its] findings and conclusions.” See 2 F.C.C.Rcd. at 5066 n. 120. After reciting and endorsing the 1985 Fairness Report’s conclusions, the Commission declared that “the fairness doctrine chills speech and is not narrowly tailored to achieve a substantial government interest.” Id. at 5057. Consequently, the FCC concluded “under existing Supreme Court precedent, as set forth in Red Lion [Broadcasting Co. v. FCC, 395 U.S. 367, 89 S.Ct. 1794, 23 L.Ed.2d 371 (1969) ] and its progeny, that the fairness doctrine contravenes the First Amendment and thereby disserves the public interest.” Id.
II.
At no time during the long and intricate proceedings in this case has any party suggested that the fairness doctrine is constitutionally compelled. Nor can it be claimed here, in view of this court’s TRAC decision, that the doctrine is statutorily mandated.1 Accordingly, the Commission has the authority to reject the doctrine if it concludes, without being arbitrary or capricious, that it no longer serves the public interest.
The Commission has slightly complicated the issue, however, by asserting that “the policy and constitutional considerations in this matter are inextricably intertwined.” 2 F.C.C.Rcd. at 5046. If that were true, we could resolve the case only by addressing the constitutional issue.
But it is an elementary canon that American courts are not to “pass upon a constitutional question ... if there is also present some other ground upon which the case may be disposed of.” See Ashwander v. Tennessee Valley Authority, 297 U.S. 288, 345-48, 56 S.Ct. 466, 482-83, 80 L.Ed. 688 (1936) (Brandéis, J., concurring). We applied the maxim in this litigation’s prior appearance here, Meredith v. FCC, 809 F.2d at 870, and it is obviously no less compelling today. Further, if an agency relies on two grounds for a decision, a court may sustain it if one is valid and if the agency would clearly have acted on that ground even if the other were unavailable. Salt River Project Agricultural Improvement and Power District v. United States, 762 F.2d 1053, 1060 n. 8 (D.C.Cir. 1985); Communication Workers of America v. NLRB, 784 F.2d 847, 851 (7th Cir.1986) (citing Bowman Transportation v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 286-87, 95 S.Ct. 438, 442, 42 L.Ed. 2d 447 (1974)).
Thus, if we are persuaded that the Commission would have found that the fairness doctrine did not serve the public interest even if it had foregone its ruminations on the constitutional issue, we must end our inquiry without reaching that issue. In fact, as we explain in part III, we have no doubt that even in the absence of constitu*42tional problems the Commission would have reached the same outcome. We then turn in part IV to petitioners’ objections to the Commission’s policy decision, and in part V to issues relating to the Commission’s withdrawal of the doctrine’s “first prong,” the requirement of coverage of controversial issues.
III.
It is quite true that the Commission at the outset of its opinion here, asserted that the constitutional and policy issues were “inextricably intertwined,” 2 F.C.C.Rcd. at 5046, and at the end suggested that its policy conclusion was a mere consequence of its constitutional one, id. at 5057 (“the fairness doctrine contravenes the First Amendment and thereby disserves the public interest”) (emphasis added). Those who would have us grasp the constitutional nettle call our attention to these references. See Intervenor Democratic National Committee Reply Brief at 4-6.
But the Commission’s reasoning behind its “intertwining” assertion belies any inference that its policy judgment depends upon its constitutional view. It first set forth two overlapping points — that “First Amendment considerations are an integral component of the public interest standard” and that “the promotion of First Amendment values was the Commission’s core policy objective in establishing and maintaining the doctrine.” Id. at 5046. Insofar as these observations state that the same values are relevant to both constitutional and policy issues, they seem unexceptionable. Surely both decisions encompass the goals of stimulating fair, balanced, and diverse treatment of controversial issues (either by single stations or by the media in the aggregate); of minimizing any chilling effect that may flow from governmental requirements that one sort of presentation be matched by another; and of minimizing the risks of abuse and other adverse effects that may flow from having governmental officials sit in judgment on editorial decisions. But it plainly does not follow from this congruity of values that the Commission can make a policy finding against the fairness doctrine only by relying on constitutional grounds.
Quite the reverse. Indeed, the Commission’s third argument for “intertwining,” namely that in resolving broadcast-related constitutional issues courts “have found our perspective informative,” see 2 F.C.C.Rcd. at 5046, pinpoints a clear distinction. In making a public interest judgment under the Communications Act,2 the Commission is exercising both its congressionally-delegated power and its expertise; it clearly enjoys broad deference on issues of both fact and policy. See, e.g., FCC v. WNCN Listeners Guild, 450 U.S. 582, 596, 101 S.Ct. 1266, 1275, 67 L.Ed.2d 521 (1981). Its role in constitutional judgments, while uncertain, is more limited. Its fact-findings have some force. Compare Columbia Broadcasting System, Inc. v. Democratic National Committee, 412 U.S. 94, 102, 93 S.Ct. 2080, 2086, 36 L.Ed.2d 772 (1973) (“[I]n evaluating the First Amendment claims of respondents, we must afford great weight to the decisions of Congress and the experience of the Commission”) with id. at 103, 93 S.Ct. at 2087 (“That is not to say we ‘defer’ to the judgment of the Congress and the Commission on a constitutional question, ... The point is, rather, that when we face a complex problem with many hard questions and few easy answers we do well to pay careful attention to how the other branches of Government have addressed the same problem”). As Judge Starr suggests, that force may depend on whether the agency has rejected or upheld the constitutional claim. See separate opinion of Judge Stair at 679 *43-680. But an ultimate constitutional decision on the doctrine necessarily melds raw facts with First Amendment value judgments, compare below at 683-684, and that melding process is for the courts.
Thus, under the legal framework within which the Commission and the courts operate, the overlap of values could not compel the Commission to rest its policy judgment on some supposition about the fairness doctrine’s constitutional status.
Of course, if the Commission had written its opinion in purely constitutional terms, we would have no choice but to address the constitutional issue or — more likely — to remand to the Commission for it to re-arrange horse and cart. We doubt that Vermont Yankee Nuclear Power Corp. v. NRDC, 435 U.S. 519, 98 S.Ct. 1197, 55 L.Ed.2d 460 (1977), would bar this court from such a remand. Compare separate opinion of Judge Starr at 677-678. While Vermont Yankee makes clear that a reviewing court is not “to impose upon the agency its own notion of which procedures are ‘best,’ ” id. at 549, 98 S.Ct. at 1214, we do not read it as saying that an agency can maneuver a court into a constitutional decision simply by vaulting over its statutory mandate to define and pursue the public interest. If courts may not resist such maneuvers, then an agency (unlike Congress) may put a hypothetical question to the courts: “If we were inclined to enforce the fairness doctrine (which we don’t think we are, but we can’t make up our minds), could we do so?”
Happily the Commission’s opinion is not written in exclusively constitutional terms. First, its intermediate conclusions, if accepted, compel a finding that the doctrine fails to serve the public interest. Most notably, in summarizing its basis for concluding that the doctrine violated the Constitution even under Red Lion’s view of the permissible scope of government control of broadcasters, it said:
In sum, the fairness doctrine in operation disserves both the public’s right to diverse sources of information and the broadcaster’s interest in free expression. Its chilling effect thwarts its intended purpose, and it results in excessive and unnecessary government intervention into the editorial processes of broadcast journalists.
2 F.C.C.Rcd. at 5052. It is hard to imagine any mental gymnastics by which the Commission, having delivered itself of such a condemnation, could go on to find the doctrine in the public interest.3
In addition, the Commission here adopted the findings of its 1985 Report wholesale. 2 F.C.C.Rcd. at 5066 n. 120. As noted above at 656, that Report never reached a conclusion on the constitutionality of the fairness doctrine, but focused explicitly and primarily on the Commission’s view that the doctrine was detrimental to the public interest. The language of this finding could not have been firmer, and was made by the Commission at three points in the 1985 Fairness Report:
On the basis of the voluminous factual record compiled in this proceeding, our experience in administering the doctrine and our general expertise in broadcast regulation, we no longer believe that the fairness doctrine, as a matter of policy, serves the public interest.
102 F.C.C.2d at 147; see also id. at 148, 246. As we observed in Meredith, the 1985 Report “largely undermined the legitimacy” and “eviscerate[d] the rationale” for the doctrine. 809 F.2d at 873. The Commission’s incorporation of the Report’s reasoning here confirms our conclusion that the Commission would have made the same public interest finding if it had approached the issue free of “intertwining.”4
*44IV.
The FCC’s decision that the fairness doctrine no longer serves the public interest is a policy judgment. There is no real dispute that fostering fair, balanced and diverse coverage of controversial issues is a good thing. Nor, so far as we can tell from the record or briefs, is there any question that discouraging any coverage at all, having government officials second-guess editorial judgments, or allowing incumbents an opportunity for abuse of power are things to be avoided, all other things being equal. The Commission’s problem was to make predictive and normative judgments about the tendency of the fairness doctrine to produce each of these things, about how bad the bad effects were and how good the good ones, and ultimately about whether bad effects outweighed good.
In this situation, we owe great deference to the Commission’s judgment. As the Supreme Court has stated,
[O]ur opinions have repeatedly emphasized that the Commission's judgment regarding how the public interest is best served is entitled to substantial judicial deference.... The Commission’s implementation of the public-interest standard, when based on a rational weighing of competing policies, is not to be set aside by the Court of Appeals, for “the weighing of policies under the ‘public interest’ standard is a task that Congress has delegated to the Commission in the first instance.”
FCC v. WNCN Listeners Guild, 450 U.S. 582, 596, 101 S.Ct. 1266, 1275, 67 L.Ed.2d 521 (1981) (quoting FCC v. National Citizens Committee for Broadcasting, 436 U.S. 775, 810, 98 S.Ct. 2096, 2119, 56 L.Ed.2d 697 (1978)).
The Commission’s factual judgments here are almost entirely predictive — statements about the overall effects of a policy on licensees and others. The Supreme Court has observed that “[I]n such cases complete factual support for the Commission's ultimate conclusions is not required since ‘a forecast of the direction in which the future public interest lies necessarily involves deductions based on the expert knowledge of the agency.’ ” FCC v. WNCN Listeners Guild, 450 U.S. at 594-95, 101 S.Ct. at 1274 (internal quotations omitted). Although we do not suppose for a minute that experience without the fairness doctrine will quell dispute — the same problems of measuring the immeasurable will remain — we bear in mind Judge Leventhal’s point that for resolution of dispute over “legislative” as opposed to “adjudicative” facts, “a month of experience will be worth a year of hearings.” American Airlines, Inc. v. CAB, 359 F.2d 624, 633 (D.C. Cir.) (en banc), cert. denied, 385 U.S. 843, 87 S.Ct. 73, 17 L.Ed.2d 75 (1966).
Of course, an agency can act arbitrarily or capriciously in the exercise of a policy judgment, and we must assure ourselves that that did not happen here. The challengers’ primary claims are that the supporting evidence was insufficient, that the Commission failed to consider the whole record or to consider alternative solutions adequately, and that its decision in the present case was an arbitrary decision from its 1985 decision to retain the doctrine.
Before addressing those specific attacks, we must describe two core findings of the 1985 Report — that the fairness doctrine often operated to chill broadcaster speech on controversial issues and that recent increases in broadcasting outlets undercut the need for the doctrine — on which the Commission relied heavily in the present decision. It found that the doctrine produced chilling effects by placing burdens on stations which chose to air numerous programs on controversial issues — including the fear of denial of license renewal *45due to fairness doctrine violations, the cost of defending fairness doctrine attacks and of providing free air time to opposing views if a fairness violation is found, and the reputational harm resulting from even a frivolous fairness challenge. While the FCC recognized that to a degree the first prong of the fairness doctrine offset this effect by requiring broadcasters to present some controversial issues, it nonetheless found that broadcasters were encouraged
to air only the minimal amount of controversial issue programming sufficient to comply with the first prong. By restricting the amount and type of controversial programming aired, a broadcaster minimizes the potentially substantial burdens associated with the second prong of the doctrine while remaining in compliance with the strict letter of its regulatory obligations. Therefore, despite the first prong obligation, in net effect the fairness doctrine often discourages the presentation of controversial issue programming.
102 F.C.C.2d at 161.
The 1985 Fairness Report also noted that paradoxically the chilling effect often fell on the expression of unorthodox or “fringe” views on controversial issues. Id. at 188. Since the doctrine compelled coverage only of “major” or “significant” opinions, the FCC claimed that in assessing fairness doctrine compliance, the Commission was called upon to evaluate broadcasters’ decisions concerning the importance of given viewpoints. The Report expressed its fear that the fairness doctrine thus had the potential “to interject the government, even unintentionally, into the position of favoring one type of opinion over another.” Id. at 190.
In assessing whether any need for the doctrine persisted, the 1985 Report found a dramatic increase in broadcasting outlets since the 1974 Fairness Report (48 F.C.C.2d 1). This expansion in broadcasting capacity was found to have been spread widely across American society. The FCC found that by 1984 96% of television households received five or more over-the-air (non-cable) television signals, compared with 837° in 1972. 102 F.C.C.2d at 205. Those receiving nine or more signals had tripled, from 217° of TV households in 1972 to 647> in 1984. Id. Of course some signals may for one reason or another be unable to function as serious alternative sources. But the new signals are plainly not trivial in the aggregate, as they have driven the networks’ audience share down from 90% in 1982 to 76% in 1984. Id. at 206.
The Commission also found significant growth in radio outlets since the 1974 Fairness Report. The number of radio stations grew by 30% between 1974 and 1985. The subset of FM service increased during the same period by 60%, leading the FCC to proclaim that “there has also been á fundamental change in the structure of the radio market. Once predominantly an AM only service, radio is now composed of two very competitive services.” Id. at 203. The Commission noted that radio expansion impacted smaller communities as well as larger urban areas; “the number of radio voices available in each local market has grown.” Id.
Looking at substitute electronic media such as low power television, video cassette recorders, satellite master antenna systems, and so forth, the Commission found that the gains in radio and television accessibility and diversity actually understated the true development of broadcast media available to the viewing public. Id. at 208-17. The Report also noted that print media coverage of controversial issues offered Americans exposure to the type of information which the fairness doctrine was designed to foster.
We now turn to the specific objections to the Commission's conclusions.
Insufficiency of Evidence to Support FCC Position
Although there is some criticism of the 1985 Report’s findings as to the growth of access to media signals, that commentary focuses largely on the point that some of the figures used by the Commission may exaggerate the significance of the alternative sources. But the Commission has included the figures on changes in audience shares, clearly reflecting its recognition *46that not all sources are equal. Some parties also assert that the FCC was overly concerned with aggregate or national figures at the expense of local data. Yet in looking at the Commission’s finding concerning the number of households which receive five or more non-cable television signals — a whopping 96% — it seems to us that only a diminutive share of the population has been unaffected by the growth the Commission has recounted. Accordingly, we view these attacks as peripheral.
Several parties, however, have attacked the evidence of broadcaster chill and what they contend is the Commission’s failure to respond adequately to the attacks. See Motor Vehicle Manufacturers Ass’n v. State Farm Mutual Automobile Ins. Co., 463 U.S. 29, 46-51, 103 S.Ct. 2856, 2868-71, 77 L.Ed.2d 443 (1983). The United Church of Christ (“UCC”) has taken the lead in this, relying in part on critical comments made in 1984 by two groups, Black Citizens for Fair Media (“BCFM”) and Media Access Project/Telecommunications Research and Action Center (“MAP/TRAC”).
In its present decision the Commission said that the record compiled for the 1985 Report contained “over 60 reported instances in which the fairness doctrine inhibited broadcasters' coverage of controversial issues.” 2 F.C.C.Rcd. at 5050. It appears that numerically the main body of evidence comes from the comments submitted to the Commission by the National Association of Broadcasters (“NAB”), which presented 45 broadcaster accounts of the effects of the fairness doctrine on their industry and policies.5 It is clear, however, that the FCC also relied on important additional sources, including submissions by some proponents of the fairness doctrine. For example, it pointed to evidence from Public Media Center that after it had warned broadcasters that it would demand free response time from stations that accepted advertising on a specific controversy, more than two-thirds of those contacted had refused to sell advertising on the subject at all. 102 F.C.C.2d at 176.
The 1985 Report responded substantially to the BCFM and MAP/TRAC attacks made on the NAB study. 102 F.C.C. 2d at 172-88. The FCC defenses took several forms. First, critics argued that the NAB examples should be heavily discounted because the participating broadcasters were “self-interested.” Id. at 180-82. The FCC noted that because “chill” is a subjective perception, broadcasters’ comments provided the best evidence of its existence: “[T]his evidence is more probative than the statements of persons who, by necessity, have to second-guess the broadcasters’ state of mind.” Id. at 181. The Commission also labeled broadcasters’ claims of chilling effect “admission[s] against interest,” as such confessions could expose broadcasters to potential fairness doctrine attacks. Id. Finally, the Commission argued that the critics’ argument proved too much: “[T]he identical charge could be leveled against every statement of every commenting party. We have never held that the evidence of interested parties lack probity; indeed, were we to adopt such a rule it would be virtually impossible for us to come to any conclusions about any issue raised in this proceeding.” Id. See Jeremy Bentham, 7 Rationale of Judicial Evidence, b.9, pt. 3, c.3, p. 393 ff. (Bowring’s ed. 1827) (“Any interest, interest of any sort and quantity, sufficient to produce mendacity? As rationale would it be to say, any horse, or dog, or flea, put to a wagon, is sufficient to move it____”) quoted in 2 Wigmore on Evidence § 576 (Chadbourn rev. 1979)). We are persuaded that the self-interested character of the broadcasters’ evidence did not bar the Commission from giving it substantial weight.
Second, the FCC responded to BCFM’s and MAP/TRAC’s allegations that many of the NAB samples were in fact evidence of *47chill resulting from the political attack rules and the political editorializing rule, rather than from the Fairness Doctrine. Id. at 186-87. The Commission observed that evidence of chilling from those rules would logically support an inference that the fairness doctrine tended to chill. Indeed, the logic seems exactly the same: if a rule defines specific conduct as a trigger for obligations, the trigger conduct will be more costly and those subject to the rule will be less inclined to indulge in it. In any event, the FCC also said that “none of the evidence of record demonstrating the existence of a 'chilling effect’ that is described in this section involves an application of the personal attack rule or the political editorializing rule.” Id. at 187 n. 157. Thus, we take it that the Commission properly focused on the evidence directly linked to the fairness doctrine itself.
Third, the Commission made specific rebuttals of some of BCFM’s and MAP/TRAC’s criticisms of NAB examples.6 For example, one of the criticisms, twice invoked, was that the particular broadcaster need not have been deterred, since in the view of the critics the station’s ordinary news coverage would have satisfied its obligations under the doctrine. In rebuttal, the Commission found the critic’s observations highly speculative. See id. at 172 n. 104 and at 173 n. 106. Surely the FCC is right. The fairness doctrine applies to ordinary mortals who adjust their affairs on the basis of estimates of risk. The estimating process costs time and effort. It hardly disproves an alleged deterrent effect to suggest-that by more careful research or analysis, or by greater brilliance, the deterred party might have come out the other way.7 This is typical of this line of critiques, and to avoid repetition we pass on to the remaining attacks.
Fourth, in the 1985 fairness proceeding, BCFM had discounted 33 of the broadcasters’ “examples” (out of 56 as recomputed by BCFM, see note 3 above) as (1) merely reflecting broadcaster hostility but not providing “concrete information” (nine cases); (2) in fact evidencing enhancement of information available to the public (six cases); (3) showing no change in broadcast behavior at all (five cases); and (4) reflecting a fairness complaint but no actual chill (11 cases).
UCC appears to be correct in saying that the FCC did not respond directly to these comments, either in the 1985 Report or the present decision. And if we thought that the Commission’s reference to “over 60 reported instances” were intended literally to refer to discrete episodes of chilling, we would be troubled by its nonresponse. In fact, however, it is clear that the Commission recognized fully that some of the NAB “examples” were not evidence of episodes but were non-episodic evidence of chilling effect. For example, while BCFM lists Example 7 in its first category — mere expressions of broadcaster hostility to the fairness doctrine, see J.A. Vol. II in Case Nos. 87-1516, 87-1544 at 510-11 — that example in fact consists of a statement by the general manager of two stations in Durango, Colorado to the effect that the pressures of the fairness doctrine cause ber stations to “think twice” about coverage of controversial issues. Id. at 424. The evidence was thus of a policy or generic response to the doctrine; the Commission expressly invoked the evidence, using NAB’s “example” label, but recognizing it as a policy not an episode. See 102 F.C.C.2d at 175 & n. 117. Thus, we think it plain that the Commission was well aware that the NAB “examples” were not all episodes. Moreover, *48as we look at the NAB “examples” and the BCFM critiques, we are not at all persuaded that the BCFM reclassifications are compelling. Example 7 is not atypical.
Finally, UCC attacks any reliance on the NAB study on the ground that it was “not based on a statistically valid sample of broadcasters’ experiences, but rather, was merely a series of anecdotal accounts.” UCC Brief at 28. The comment is quite valid, but in the absence of either any statistically valid evidence on the other side, or even a suggestion of how the Commission could have constructed a statistically valid study, we are perplexed as to what the Commission was supposed to have done. Editorial decisions are obviously driven by many factors. Isolation of causes in any scientific way seems virtually impossible. The fairness doctrine has been applicable in one form or another from 1949 until the present decision, so the Commission could not compare stations' practices under the rule with their conduct free of the rule. (Comparison to practice in other nations would encounter the usual cross-cultural difficulties.)
We note that when speaking of a state-enforced “right:of-reply” applicable to newspapers, the Supreme Court has taken it as self-evident that such a duty “inescapably ‘dampens the vigor and limits the variety of public debate.’ ” Miami-Herald Publishing Co. v. Tornillo, 418 U.S. 241, 257, 94 S.Ct. 2831, 2839, 41 L.Ed.2d 730 (1974) (emphasis added). See also Pacific Gas & Electric Co. v. Public Utilities Comm’n of California, 475 U.S. 1, 106 S.Ct. 903, 908, 89 L.Ed.2d 1 (1986) (citing Tomillo to support assumption that California’s “[cjompelled access” to public utility billing envelopes “penalizes the expression of particular points of view”). Of course we recognize that the. fairness doctrine differs from the right-of-reply at issue in Tomillo and that Red Lion applies different constitutional standards to the broadcast media. But we think the Court’s approach suggests that where a rule imposes potentially onerous and at least irksome consequences on the exercise of speech, there is nothing very startling about an inference that the rule will often deter-speech. We think the Commission could properly find that effect from the evidence before it here.
Failure to Consider Evidence of Doctrine’s Beneficent Effects
UCC also charges that the Commission ignored or discounted evidence of the benefits derived from the fairness doctrine. See UCC Brief at 41-46. We think part of the problem here arises out of UCC’s apparent assumption that the Commission believed that the fairness doctrine rarely (or never) increased diversity of expression. Indeed, the Commission only purported to find that “the overall net effect of the doctrine is to reduce the coverage of controversial issues of public importance,” 2 F.C.C.Rcd. at 5050, a finding clearly consistent with a belief that the doctrine frequently produced its intended effects. We do not read the Commission’s focus on the negative aspects as manifesting a blindness to unwelcome evidence but rather as a focus on what it viewed as novel or surprising. To say that a rule has often produced its intended effects is to tell a dog-bites-man story — not front-page stuff. The Commission was naturally more struck by evidence of ««intended consequences— hardly in the man-bites-dog class, but closer.
In one instance, for example, UCC claims that the FCC twisted some comments which illustrated the efficacy of the doctrine and attempted to use this evidence to show that broadcasters were being chilled as a result of public interest groups’ pressure. It points to the Commission’s treatment of evidence submitted by the Public Media Center describing an anti-nuclear coalition’s campaign to secure free time by invocation of the doctrine. The coalition had indeed asked for and received free time, but it also had invited stations to refuse to sell time and thereby to avoid any fairness problem at all. Over two-thirds pursued that course, whether because of the doctrine or for other reasons. The Commission said this “vividly illustrates the manner in which a complainant can successfully pressure broadcasters into refusing to sell advertising on ballot issues.” *49102 F.C.C.2d at 176. In the context of this proceeding, we do not read the Commission’s characterization as reflecting blindness to the doctrine’s positive effects.8
We must confess, however, some perplexity at the Commission’s insistence that the doctrine’s overall net effect [was] to reduce the coverage of controversial issues.” 2 F.C.C.Rcd. at 5050. The Commission was plainly driven by its effort to resolve the issue not only in policy terms but also in constitutional ones, and it expressly characterized the “net effect” as being “in contravention of the standard announced in Red Lion.’’ Id. The latter indeed said that if in the future it were shown that the fairness and kindred doctrines had “the net effects of reducing rather than enhancing the volume and quality of coverage, there will be time enough to reconsider the constitutional implications.” 395 U.S. at 393, 89 S.Ct. at 1808. The Court then proceeded to declare, “[T]he fairness doctrine in the past has had no such overall effect.” Id. The Court did not explicate the basis for that finding.
We are frankly uncertain how anyone could be sure either way: The definitional problems alone are staggering. How could a court or agency persuasively define standards by which to evaluate the “quality” of coverage? Once it had done so, by what yardstick could it balance quality increments due to broadcasters’ responding as intended against quality decrements due to the deterrent effects? The Commission does not explain. Nor do the challengers, who evidently are happy to weigh these unweighables but read the scales differently-
Despite this confusion, we believe we can “discern the path” followed by the Commission. See Bowman Transportation, 419 U.S. at 286-89, 95 S.Ct. at 442-43. It found itself confronted with evidence that it read as establishing that the doctrine has very substantial deterrent effects (and, as we have said, we believe that finding was permissible). It also quite plainly made the normative judgment that government decisions on program content and government-forced expressions of ideas, together with the associated risks of partisan abuse, were anathema. See 102 F.C.C.2d at 190-94; 2 F.C.C.Rcd. at 5050-51, 5055-57. It explicitly found support for that value judgment in Supreme Court analyses of the First Amendment as applied to parties other than broadcasters, citing Miami-Herald Publishing Co. v. Tomillo and Pacific Gas & Elec. Co. v. Public Utilities Comm 'n of California. See 2 F.C.C.Rcd. at 5057. It was persuaded that the doctrine’s deterrent effects were on the same scale as its expression-generating ones, and it plainly discounted the value of the latter as “governmentally-coerced speech.” Id. Accordingly it rejected the doctrine as a matter of policy. That policy judgment seems to us by no means arbitrary or capricious, and indeed quite similar to the Supreme Court’s finding of similar value judgments implicit in the First Amendment as applied to the print media. And the Commission's course is not rendered arbitrary or capricious by its choosing, as the Court had in Red Lion, to cast its decision in terms of an assessment of the “net effect” of the doctrine.
Failure to Consider Alternatives
Petitioners Geller and Lampert contest the FCC’s refusal to adopt their proposed alternative to the elimination of the fairness doctrine. The two have repeatedly urged the Commission to review fairness matters only at license renewal, and to adopt a malice standard for this review.
The FCC has twice declined to return to its pre-1962 practice of resolving fairness doctrine complaints at renewal, first in its 1974 Fairness Report, 48 F.C.C.2d at 17-18, affirmed by this court in National Citizens *50Committee for Broadcasting v. FCC, 567 F.2d 1095 (D.C.Cir.1977), cert. denied, 436 U.S. 926, 98 S.Ct. 2820, 56 L.Ed.2d 769 (1978), and then in its 1987 Alternatives Report, 2 F.C.C.Rcd. 5272, 5276-78. The 1987 Report found that
although the renewal-only proposals might appear, at first glance, to lessen the degree of immediate, direct and tangible government intrusion into the editorial processes of broadcast licensees, in actuality the net effect would be merely to postpone the eventuality of extensive governmental involvement at license renewal time.
Id. at 5278. The Commission noted several specific defects that it perceived in the renewal review plan. First, by linking fairness evaluations with the renewal process, broadcasters would view the potential sanctions as being more threatening than those currently in effect, as loss of license or suspension would be seen as more likely. Id. Second, the renewal plan deprives the broadcaster of a contemporaneous ruling on its behavior and chance to quickly rectify any imbalance in its programming. The FCC believed this could make broadcasters more conservative in coverage of controversial issues. Id. Finally, it would only delay rather than diminish the government’s meddling in the editorial process. Id.
Geller and Lampert coupled their renewal-only suggestion with the idea that review would proceed under a “malice” standard such as that of New York Times Co. v. Sullivan, 376 U.S. 254, 84 S.Ct. 710, 11 L.Ed.2d 686 (1964), and with an alternative suggestion that the Commission continue to act case-by-case but under a malice standard. As to the first, the Commission agreed that such a standard would reduce the chilling effect, but judged that reduction inadequate to offset what it believed would be the increased chill from postponement to renewal. 1987 Alternatives Report, 2 F.C.C.Rcd. at 5278. As to the second, the Commission in its reconsideration order in the present case expressed doubt whether a malice standard would substantially reduce the chilling effect when compared with the Commission’s practice of deferring to broadcasters’ reasonable judgments, 3 F.C.C.Rcd. at 2037, and said that the difference in impact on licensees would be a matter of degree rather than of kind, id. at 2038. Although its discussion of the subject is more than usually framed in constitutional terms, we read it as concluding that the proposals only mildly mitigated the ill effects that drove the Commission to reject the fairness doctrine as previously enforced. That judgment seems well within its discretion, especially as the proposals would also appear to materially diminish the positive effects of the doctrine.
Departure from 1985 Decision
Some parties complain that the present decision represents an unexplained abandonment of the Commission’s 1985 decision to persist in enforcement of the doctrine. Cf. Motor Vehicle Mfrs. Ass’n v. State Farm Mutual Automobile Ins. Co., 463 U.S. 29, 41-44, 103 S.Ct. 2856, 2865-67, 77 L.Ed.2d 443 (1983) (agency changing course required to supply reasoned analysis for the change); Center for Auto Safety v. Peck, 751 F.2d 1336, 1343 (D.C.Cir.1985).
The 1985 Fairness Report had found that the doctrine disserved the public interest, but had refrained from dropping the doctrine on policy grounds only because of concern that perhaps it was statutorily mandated, 102 F.C.C.2d at 227, and because of the intense congressional interest then brewing over the fairness doctrine, id. at 246-47. This court’s TRAC decision removed the first obstacle (except insofar as the Supreme Court might find otherwise); a successful presidential veto of Congress’s attempt to mandate the fairness doctrine clearly diminished the second. Though we find no explicit reference to these reasons in the decision under review, they appear so obvious and compelling that a remand to extract the magic words from the Commission would be pure waste. See Bowman Transportation, 419 U.S. at 286-89, 95 S.Ct. at 442-43.
V.
Several parties attack the Commission’s repeal of the so-called “first prong” of the *51fairness doctrine — the requirement that broadcasters provide coverage of important controversial issues of interest to the community they serve. First, they contend that they were not adequately notified of the Commission’s intent to reconsider the first prong in this proceeding. Second, petitioners argue that the FCC’s abandonment of the first prong was arbitrary and capricious.
Notice
We must first explain the relevance of notice. The decision under review has been, up to this point, an adjudication in the classic sense. The Commission disposed of a complaint against Meredith Corporation; its rationale for that decision was that the doctrine under which Meredith had been sanctioned did not serve the public interest (and was unconstitutional). Up to this point we have reviewed the decision as we would any agency adjudication.
The Commission’s rejection of the first prong of the fairness doctrine, however, was surely not necessary to its exoneration of Meredith here. No one had ever suggested that Meredith had failed to cover controversial issues; quite the contrary, the whole point was that its coverage was enough to trigger a duty under the second prong. Accordingly, viewing the matter purely as an adjudication, the Commission’s statements on the first prong might well be dictum and on that account probably unreviewable. See Office of the Consumers’ Counsel v. FERC, 808 F.2d 125, 128-29 (D.C.Cir.1987).
If, however, the Commission complied with the procedural requirements for an informal rulemaking under 5 U.S.C. § 553 (1982), and if it spoke in the dispositive terms that distinguish a substantive rule from a mere policy statement, see Vietnam Veterans of America v. Secretary of the Navy, 843 F.2d 528, 535-39 (D.C.Cir.1988), then its treatment of the first prong constitutes a substantive rule binding the agency and clearly reviewable. There can be no question here that the Commission purported to dispatch the first prong of the doctrine definitively, and the only procedural flaw that opponents assert is the want of notice.
It seems to us clear that the relevant notices reasonably warned interested parties that the Commission was looking at the fairness doctrine as a whole. The Notice of Inquiry leading to the 1985 Fairness Report, 49 Fed.Reg. 20317 (May 14, 1984), said:
Our purpose in instituting this inquiry is to undertake the most searching and comprehensive reexamination of the fairness doctrine that this agency has ever had. Our goal is to obtain as complete a record as possible by inviting comment on all facets of the fairness doctrine ...
Id. at 20318 (emphasis added).
On remand from our decision in Meredith the Commission issued a notice telling interested parties of its plan to review the constitutional and public interest underpinnings of the “fairness doctrine.” While that notice did not employ the especially sweeping terms used in the Notice of Inquiry, it spoke of the fairness doctrine generally and specifically linked the proceeding to the prior Inquiry:
[tjhe Commission hereby requests comment on whether, in light of the 1985 Fairness Report, enforcement of the fairness doctrine is constitutional and whether enforcement of the doctrine is contrary to the public interest. Although the Commission does not ordinarily seek public comment in an adjudicatory proceeding, because of the general importance of the issues in this particular case, the Commission will entertain comments from all interested members of the public
Order Requesting Comment, FCC 87-33 (Jan. 23, 1987), 52 Fed.Reg. 2805-01 (Jan. 27, 1987).
We believe these provided adequate notice that both aspects were under review.
Whether Rescission of the First Prong Was Arbitrary and Capricious
In explaining its rejection of the first prong, the Commission has argued that the doctrine is a unified whole, so that, once it *52rejected the second requirement on policy and constitutional grounds, it would be inappropriate to preserve the first. 2 F.C.C. Red. at 5048. We regard this as a statement of a conclusion, not a reason, so that if it stood alone we would be compelled to reverse this aspect of the Commission’s action. The Commission analogizes its action to that of a court that has held part of a statute unconstitutional and must then consider whether to sever other parts. Reconsideration Order, 3 F.C.C.Rcd. at 2038. The analogy only underscores the deficiency of the Commission’s “unified doctrine” declaration. In making a severability decision a court essentially inquires what the legislature would have done had it been aware of the infirmity of one part. Here, the Commission was acting both as adjudicator and legislative body, and was thus perfectly free to decide what it wanted. Moreover, even a court making a severability decision tries to explain why it believes the legislature would prefer the valid portions of the statute to fall or survive. See, for example, Bowsher v. Synar, 478 U.S. 714, 106 S.Ct. 3181, 3192-93, 92 L.Ed.2d 583 (1986).
In other portions of its decision, however, the Commission went on to supply reasons for terminating the first prong. The Commission’s vigor in expressing these reasons leaves us no serious doubt that it would have proceeded to end the first prong even if it had recognized that its “unified doctrine” statement was a mere conclusion.
First, removal of the fairness doctrine’s second requirement would reduce the need for the coverage requirement. With the chilling effects of the second requirement ended, the Commission expected that “coverage of controversial issues will be forthcoming naturally, without the need for continued enforcement of the first prong.” 3 F.C.C.Rcd. at 2038; see also 102 F.C.C.2d at 161 n. 66. It viewed the first prong as having originated in part as a “backstop[ ]” to the second prong, so that removal of the latter removed much of the reason for the former. 3 F.C.C.Rcd. at 2038.
Second, it viewed the coverage requirement as in significant part duplicative of its independent requirement that broadcasters cover issues “of importance” to their communities. 2 F.C.C.Rcd. at 5048.9 While the FCC acknowledged on reconsideration that the two programming requirements were not identical, it saw sufficient similarity to believe that the community issues rule would fill any material regulatory gap. 3 F.C.C.Rcd. at 2039.
Third, in its discussion of the fairness doctrine as a whole the Commission relied heavily on its view that government involvement in the editorial process was offensive. See 102 F.C.C.2d at 190-94; 2 F.C.C.Rcd. at 5050-52, 5055-57; above at 26. That judgment of course applies to the editorial decisions required for enforcement of the first prong, and the Commission made the point expressly: “[T]he doctrine requires the government to second-guess broadcasters’ judgment on such sensitive and subjective matters as the ‘controversiality’ and ‘public importance’ of a particular issue____” 2 F.C.C.Rcd. at 5052. It also alluded to the offensive character of government first-prong decisions in distinguishing that part of the doctrine from the community issues requirement:
While enforcement of the doctrine’s first prong requires the government to judge, on a case-by-case basis, whether a specific issue is both controversial and of vital importance to mandate coverage by the broadcaster, enforcement of the issue responsive obligation requires a different level of government intervention in determining, at renewal time, whether broadcasters’ overall programming covered the needs and interests of its community.
3 F.C.C.Rcd. at 2042 n. 67 (emphasis in original). Compare separate opinion of Judge Wald at 671-72 (suggesting that *53the Commission failed to identify any drawbacks in enforcement of the first prong).
We believe that these reasons, particularly in light of the Commission’s background findings on the increased diversity of outlets and programming, adequately support its removal of the first prong.
Conclusion
We conclude that the FCC’s decision that the fairness doctrine no longer served the public interest was neither arbitrary, capricious nor an abuse of discretion, and are convinced that it would have acted on that finding to terminate the doctrine even in the absence of its belief that the doctrine was no longer constitutional. Accordingly we uphold the Commission without reaching the constitutional issues. The petition for review is denied.
. Thus the claim of some parties that the Communications Act incorporates a public trustee concept that necessarily includes the fairness doctrine is in essence an effort to ask this panel to overturn TRAC.
. The Commission imposes a continuing obligation on licensees to operate "in the public interest" via its power under 47 U.S.C. § 312(b) to issue cease-and-desist orders for failing to operate substantially as set forth in their licenses. The Supreme Court has made clear that the Commission's authority to require continued operation "in the public interest" may be either explicit in the conditions of license issuance or implicit in the FCC’s authority under 47 U.S.C. § 312(a)(2) to revoke a license for conditions that would justify denial of an initial license, coupled with the statutory requirement that the public interest be served in granting and renewing licenses, id. at §§ 307(a) and (d). See Red Lion, 395 U.S. at 372 n. 3, 89 S.Ct. at 1797 n. 3.
. In his separate opinion, Judge Starr suggests at 677 that the Commission has "view[ed] the public interest as driven by First Amendment values (when other[s] ... were available)____’’ The Commission’s adoption of that view (if that's what it is) does not seem to us to automatically transform a public interest decision based on such variables into a constitutional decision.
. Even if all parties chose to frame the issue as exclusively constitutional, see separate opinion of Judge Starr at 13, these litigation strategies would not establish that the Commission had rested its decision exclusively on constitutional grounds. In any event, while it is true that most parties here seek either to condemn or to vindicate the doctrine on constitutional grounds, not *44all do. Petitioner Office of Communication of the United Church of Christ squarely attacks the decision as "arbitrary and capricious.” See its opening brief at 2, 10, 11, 40, 46. Moreover, in extended passages of their briefs the parties have been content to discuss the Commission's intermediate conclusions, such as those relating to "chill" or the growth of broadcasting outlets, without confining their analysis to constitutional (as opposed to policy) criteria. See, e.g., Brief of Intervenor Meredith Corporation at 17-22 (passage captioned, "The Explosive Growth of Communications Media Renders the Doctrine Unnecessary”).
. The exact number of examples in the NAB study is subject to dispute, as the BCFM Comments indicate. That group claimed that the NAB study had some accounts which actually incorporated more than one example, while others which were numbered separately actually referred to the same underlying event. BCFM asserted that, when these discrepancies were accounted for, there were 56 examples. BCFM Reply Comments at 55-56, J.A. Vol. II in Case Nos. 87-1516, 87-1544 at 509-10.
. Id. at 172-73 nn. 104-07; id. at 174 n. 109.
. It might support an alternative doctrine with more clearly chiselled edges, but no party claims to have presented that alternative in the administrative proceedings. UCC’s Brief at 37 n. 43 notes the existence of a draft FCC "primer” on the doctrine, prepared in 1981 but never issued, and suggests that its issuance could have been "highly effective at eliminating misperceptions about the ... doctrine." Having examined the 51-page, singled-spaced document, J.A. (in Nos. 87-1516, 87-1544) at 145-96, we think it far from self-evident that its issuance would have drastically reduced the penumbral chill of the doctrine. Under no circumstances, of course, could it eliminate the non-penumbral chill, i.e., the doctrine’s disincentive effects on those who grasped its meaning with perfect accuracy.
. UCC claims that the Commission "completely ignored" comments submitted by the Safe Energy Communications Council giving 25 examples of successful uses of the doctrine. UCC Brief at 42. This is flatly untrue. The Commission expressly considered the Council’s comments, 2 F.C.C.Rcd. at 5066 n. 139, stating that in light of the evidence gathered in the 1985 Report the Commission remained "convinced that the net effect of the doctrine [was] to inhibit the expression of opinion on important public issues.”
. The community issues requirement is set out in Deregulation of Radio, 84 F.C.C.2d 968, 983 n. 36, recon. denied, 87 F.C.C.2d 797 (1981), rev'd in pert on other grounds, Office of Communication of United Church of Christ v. FCC, 707 F.2d 1413, 1430-32 (D.C.Cir.1983); Commercial TV Stations, 98 F.C.C.2d 1076, 1092 n. 54 (1984), recon. denied, 104 F.C.C.2d 358 (1986), rev'd in part on other grounds, Action for Children’s Television v. FCC, 821 F.2d 741 (D.C.Cir.1987).