Henry Geller, with whom Donna Lampert and Andrew Schwartzman, were on the brief, for intervenor, United Church of Christ.

Michael S. Horne, with whom J. Laurent Scharff, James M. Smith and Henry L. Baumann, were on the brief, for intervenors, Nat. Ass'n of Broadcasters and Ass'n of Independent Television Stations, Inc. Julian L. Shepard and Molly Pauker also entered appearances for intervenor, National Ass'n of Broadcasters.

Arthur Pankopf, Susan Dillon, Baryn S. Futa, Martha M. Zornow, Paula A. Jameson and Nancy H. Hendry, were on the brief, for intervenors, Corp. for Public Broadcasting, et al.

Norman P. Leventhal, Raul R. Rodriquez, Sally A. Buckman, and Richard F. Swift, were on the joint brief for intervenors, Spanish Intern. Communications Corp., Nat. Independent Television Committee and Univision, Inc. Judith Whittaker, was on the brief, for intervenor, Spanish Intern. Communications Corp. Richard E. Wiley and John C. Quale, also entered appearances for intervenor, Spanish Intern. Communications Corp.

Michael D. Berg, entered an appearance for intervenor, Lincoln Broadcasting Co.

Jay E. Ricks, entered an appearance for intervenor, Nat. Cable Television Ass'n, Inc., et al.

Before WALD, Chief Judge, and MIKVA, Circuit Judge and McGOWAN, Senior Circuit Judge.

Opinion for the Court filed by Chief Judge WALD.

Two years ago, in Quincy Cable TV, Inc. v. Federal Communications Commission, 768 F.2d 1434 (D.C.Cir.1985), cert. denied sub. nom. National Association of Broadcasters v. Quincy Cable TV, Inc., 476 U.S. 1169, 106 S. Ct. 2889, 90 L. Ed. 2d 977 (1986), we struck down as violative of the first amendment the FCC's "must-carry" rules. Those rules required cable television operators, upon request and within the limits of their channel capacity, to transmit to their subscribers every over-the-air television broadcast signal that was "significantly viewed in the community" or otherwise considered "local" under the Commission's rules. See Quincy Cable TV, 768 F.2d at 1437. Today, we revisit this distinctive corner of first amendment jurisprudence, to evaluate the constitutional validity of the scaled-down must-carry rules adopted by the FCC following our decision in Quincy Cable TV. Although the FCC has eliminated the more extreme demands of its initial set of regulations, its arguments in this case leave us unconvinced that the new must-carry rules are necessary to advance any substantial governmental interest, so as to justify an incidental infringement of speech under the test set forth in United States v. O'Brien, 391 U.S. 367, 88 S. Ct. 1673, 20 L. Ed. 2d 672 (1968). Accordingly, we invalidate as incompatible with the first amendment this latest incarnation of the FCC's must-carry rules.

Since the mid-1960's, when the nascent cable television industry began to loom as a threat to ordinary broadcast television, the Federal Communications Commission has labored to protect the local broadcast media through regulation of the cable industry. The Commission's objective in these endeavors

was not merely to protect an established industry from the encroachment of an upstart young competitor, although such a result was clearly the byproduct of the regulatory posture that developed. Rather, the Commission took the position that without the power to regulate cable it could not discharge its statutory obligation to provide for "fair, efficient, and equitable" distribution of service among "the several States and communities." If permitted to grow unfettered, the Commission feared, cable might well supplant ordinary broadcast television. A necessary consequence of such displacement would be to undermine the FCC's mandate to allocate the broadcast spectrum in a manner that best served the public interest. In particular, if an unregulated, unlicensed cable industry were to threaten the economic viability of broadcast television, the Commission would be powerless to effect what it saw (and continues to see) as one of its cardinal objectives: the development of a "system of [free] local broadcasting stations, such that 'all communities of appreciable size [will] have at least one television station as an outlet for local self-expression.' "

Quincy Cable TV, 768 F.2d at 1439 (citations and footnote omitted). See also United States v. Southwestern Cable Co., 392 U.S. 157, 88 S. Ct. 1994, 20 L. Ed. 2d 1001 (1968) (approving FCC regulation of cable as within the agency's authority so long as its actions are "reasonably ancillary" to its regulation of broadcast television); Amendment of Part 76 of the Commission's Rules Concerning Carriage of Television Broadcast Signals by Cable Television Systems, 1 F.C.C. Rcd 864 (1986) (hereinafter "Report and Order "), reconsid. denied, 2 F.C.C. Rcd 3593 (hereinafter, "Recon. Order "), at paragraphs 1-29 (tracing history of cable regulation).1

Must-carry rules in various forms have been major tools in this campaign to protect local broadcasting from cable. The FCC first introduced such rules in 1962, when it sought to impose a must-carry requirement as a condition for granting an application to construct a microwave system to transmit distant signals to a rural cable system. See Carter Mountain Transmission Corp., 32 F.C.C. 459 (1962), aff'd, 321 F.2d 359 (D.C.Cir.), cert. denied, 357 U.S. 951, 84 S. Ct. 442, 11 L. Ed. 2d 312 (1963); see also Quincy Cable TV, 768 F.2d at 1440 n. 11. In time, the FCC developed a broader must-carry regime, generally requiring cable operators, "upon request, to carry any broadcast signal considered local under the Commission's complex formula." Quincy Cable TV, 768 F.2d at 1440. The philosophy behind these rules was

to assure that the advent of cable technology not undermine the financial viability of free, community-oriented television. If cable were to "drive out television broadcasting service ... the public as a whole would lose far more--in free service, in service to outlying areas, and in local service to outlying areas, and in local service with local control and selection of programs--than it would gain." The must-carry rules, together with a comprehensive body of related regulations, would channel the development of the nascent cable industry to limit the risks it might pose to conventional broadcasting, "society's chosen instrument for the provisions of video services."

Id. (citations omitted); see generally id. at 1440-43 (describing, in considerably greater detail, the rationale for the pre-Quincy Cable TV must-carry rules).

In 1985, this circuit faced for the first time the question whether the broad must-carry rules which had been in existence for nearly two decades were in harmony with the first amendment. Judge Wright's opinion for a unanimous panel in Quincy Cable TV held that they were not. As a threshold matter, we observed that our first amendment review of regulations burdening cable television was not governed by those cases, such as Red Lion Broadcasting Co. v. FCC, 395 U.S. 367, 89 S. Ct. 1794, 23 L. Ed. 2d 371 (1969) and FCC v. League of Women Voters of California, 468 U.S. 364, 104 S. Ct. 3106, 82 L. Ed. 2d 278 (1984), upholding regulations on broadcast television. In reaching that conclusion, we noted "the Supreme Court's oft-repeated suggestion that the First Amendment tolerates far more intrusive regulation of broadcasters than of other media precisely because of the inescapable physical limitations on the number of voices that can simultaneously be carried over the electromagnetic spectrum." 768 F.2d at 1448. Wire-carried media like cable, of course, have no such limitations, and thus we found the "scarcity rationale" that the Supreme Court has used to justify broadcast television regulations to offer no succor to those seeking to establish the constitutional validity of cable television regulations. Id. at 1448-50.

Quincy Cable TV did not, however, establish the precise degree of first amendment protection enjoyed by cable operators. Although our opinion noted that some parallels existed between the must-carry regulations and regulations impinging on editorial discretion that had been invalidated in the past, see Quincy Cable TV, 768 F.2d at 1452 (citing Miami Herald Co. v. Tornillo, 418 U.S. 241, 94 S. Ct. 2831, 41 L. Ed. 2d 730 (1974)), it pointedly declined to "definitively decide" whether cable operators enjoy the heightened protection accruing to newspapers or whether the must-carry regulations were more appropriately evaluated under the test, set forth in United States v. O'Brien, 391 U.S. 367, 88 S. Ct. 1673, 20 L. Ed. 2d 672 (1968). See Quincy Cable TV, 768 F.2d at 1454. Rather, we concluded that the must-carry rules would fail even the O'Brien test's requirement of a substantial governmental interest furthered by means no greater than are essential to the furtherance of that interest.

The reasons for our invalidation of the 1985 must-carry rules under the O'Brien test were twofold. First, we concluded that the Commission had not adequately substantiated its assertion that a substantial governmental interest existed. In Quincy Cable TV we stated that, even accepting the view that the preservation of free local television was an important regulatory goal, our review of the FCC's reports and regulations suggested that the problem the sweeping must-carry rules purported to prevent--the destruction of free, local television--was merely a "fanciful threat," unsubstantiated by the record or by two decades of experience with cable TV. Id. at 1457. In general, we noted, "the mere abstract assertion of a substantial governmental interest, standing alone, is insufficient to justify the subordination of First Amendment freedoms." Id. at 1454. Second, even if the interest had been deemed substantial, the broadly-drafted must-carry rules represented a fatally overinclusive response to the problem. We observed in this vein that the rules indiscriminately protected every local broadcaster, regardless of whether it was in fact threatened, and regardless of the quantity of local service available in the community and the degree to which the cable operator in question already carried local outlets. Id. at 1459-62. We did, however, note that our decision in no way foreclosed the Commission from adopting new must-carry rules consonant with the O'Brien test. Id. at 1463.

In the aftermath of Quincy Cable TV, the FCC immediately suspended enforcement of the must-carry rules. Four months later, it announced its intention to undertake rulemaking proceedings, see Notice of Inquiry and Notice of Proposed Rulemaking, 50 Fed.Reg. 48232 (1985), and eventually, in November 1986, 16 months after Quincy Cable TV had been handed down, the agency released a new, more limited set of must-carry rules designed to accommodate Quincy Cable TV's concerns. See Report & Order. In the decision to promulgate these new rules, the Commission took note of the many comments, submitted primarily but not exclusively by broadcasting interests, arguing that some form of FCC intervention remained necessary to protect local broadcasting. See Report & Order at paragraphs 36-51; see also id. at paragraphs 52-57 (describing comments, primarily from cable operators, arguing that the reinstitution of must-carry was unnecessary and undesirable).

The most salient feature of the new rules was that the Commission substantially altered its stated justification for imposing must-carry rules at all. No longer did the Commission argue, as it had prior to the Quincy Cable TV decision, that the rules were needed for the indefinite future to ensure viewer access to local broadcast stations. Rather, the Commission now argued that must-carry rules were needed to guarantee such access during a shorter-term transition period during which viewers could become accustomed to an existing and inexpensive but largely unknown piece of equipment known as the "input-selector device."

Such devices, if hooked up to a television, allow viewers at any given time to select, simply by flicking a switch, between shows offered by their cable system and broadcast television shows offered off-the-air. These devices, the most common of which is known in the cable industry as an "A/B switch," are about the size of a standard lightswitch, and work by being hooked up to a roof-top, attic or television-top antenna. According to a study cited by the Commission in its report explaining the new must-carry rules, the cost of buying such a switch is approximately $7.50, and the cost of buying an outdoor antenna to go with it is approximately $50. See Joint Appendix ("J.A.") at 240-42 (cited at Report & Order at p 124). Outdoor antennas are generally the more expensive of the three types of antennas.

The Commission estimated that it would take approximately five years for the public to become acclimated to the existence of these switches, and accordingly, its interim rules should be in place for that same five years. See 47 C.F.R. Sec. 76.64 (stating that rules remain in force until January 15, 1992); see also Report & Order at p 138. At that point, the need for ongoing must-carry rules to ensure viewer access to local broadcast stations would be obviated. See Report & Order at p 163 ("once cable subscribers become accustomed to using off-the-air reception on an equal basis with cable service, then cable systems no longer will have an artificial ability to limit their subscribers' access to over-the-air broadcast signals"); see also id. at p 138 ("While we have found that short-term must carry regulations are necessary in order to ensure that broadcasting remains a competitive alternative source of programming in the interim period, the record clearly supports no more extensive regulatory program than that which we are adopting").

Because the Commission envisioned these switches as guaranteeing effective viewer choice between local and cable shows, it ultimately added to the new must-carry regime the requirement that cable systems offer subscribers, for pay, input-selector devices that could be hooked up to their TVs. See 47 C.F.R. Sec. 76.66; see also Report & Order at p 140; Recon. Order at paragraphs 80-94 (sketching input-selector requirements and amending earlier regulations so as not to require cable operators to install such devices for free or at cost). It did so over the reservations of some broadcasting concerns, who viewed the input-selector devices as less protective than must-carry rules. See Report & Order at paragraphs 45-47 (noting that "broadcasting interests" did not regard the A/B switch as an efficacious way of protecting local broadcasting). The Commission, observing that relatively few consumers knew about the switch-and-antenna mechanism and noting that the long history of must-carry rules had created a public "misperception" that "broadcast signals will always be available as part of their basic cable service," see Report & Order at paragraphs 121-22, also promised to require cable operators to educate the viewing public about the availability of the switch-and-antenna mechanism. See, e.g., Report & Order at paragraphs 1, 136.

In addition to thus offering a new and more limited justification for must-carry rules, the Commission also substantially limited the sweep of the new rules in a number of respects. It set forth limits on how many channels a cable carrier must devote to must-carry: carriers with 20 channels or less were not required to carry any must-carry stations; carriers with between 21 and 26 stations could be required to carry up to 7 channels of must-carry signals; and carriers with 27 or more channels could be required to devote up to 25% of their system to must-carry signals. See 47 C.F.R. Sec. 76.56; see also Report & Order at paragraphs 150-52. It also limited the pool of potential must-carry channels to those satisfying a "viewing standard" generally demonstrating a minimum viewership of the channel in question. See 47 C.F.R. Sec. 76.5(d)1(ii); 47 C.F.R. Sec. 76.55 (stating that a broadcast station qualifies for inclusion in must-carry pool if it demonstrates that it attains at least an average share of total viewing hours of at least 2 percent and a net weekly circulation of 5 percent in noncable households in the county where the cable system is located); see also Report & Order at paragraphs 145-46. The Commission also authorized cable operators to refuse to carry more than one station affiliated with the same commercial network. See Report & Order at p 153. Finally, the Commission limited the number of noncommercial stations required to be carried, stating that when the cable system had fewer than 54 channels and an eligible noncommercial station or translater existed, the cable operator must devote at least one channel to a noncommercial station; and that when the cable system had 54 or more stations, it must devote two must-carry channels to such endeavors. See 47 C.F.R. Sec. 76.56.

Constitutional and statutory challenges to these new must-carry rules were lodged shortly after their promulgation by an array of cable operators and public interest group. Petitioner Century Communications Corp., joined by 13 other cable operators (hereinafter "Joint Petitioners"), protests the must-carry rules as violative of the first amendment of the Constitution, as a taking of property without just compensation in violation of the fifth amendment, and as a measure not authorized by the FCC's statutory jurisdiction and hence ultra vires. Petitioner Richard Leghorn, a former cable system operator and presently an investor in the cable industry, challenges the rules on first amendment grounds. Petitioner Hubbard Broadcasting, Inc., a broadcasting concern, assails the failure of the new rules to afford must-carry rights to commercial broadcast translator stations as arbitrary and capricious and therefore violative of the Administrative Procedure Act ("APA"). Intervenor National Independent Television Committee argues that the must-carry rules are inconsistent with the Commission's statutory charter to protect adequately needy local stations. Intervenor the United Church of Christ challenges the regulations as arbitrary and capricious in a number of respects. Three other intervenors, the National Independent Television Committee, Spanish International Communications Corp., and Univision, Inc., target the viewing standard provision of the new must-carry rules as a content-based regulation giving preference to "popular" over "unpopular" speech and therefore violative of the first amendment; these groups also contend that this requirement is an arbitrary and capricious measure adopted in violation of the APA.

The FCC, in response, defends the must-carry rules as based on a satisfactory administrative record and as consonant with the first and fifth amendments. In this endeavor it is joined by five intervenors. The Corporation for Public Broadcasting, the National Association of Public Television Stations, and the Public Broadcasting Service defend the FCC initiative as consistent with both the first amendment and the APA. Two other intervenors, the National Association of Broadcasters and the Association of Independent Television Stations, Inc., similarly defend the regulations against constitutional and statutory attack.

We, however, need look no further than petitioners' first amendment claims to decide this case. Because we invalidate the entire new must-carry regime as unjustified and as unduly sweeping, we do not reach--and therefore express no opinion on--the subsidiary first amendment challenges to particular facets of the rules, or the arguments based on the APA that the rules are too narrow in scope.

A. The Appropriate Level of First Amendment Scrutiny

A threshold question for our first amendment analysis is what standard of review to apply. As in Quincy Cable TV, the parties dwell heavily on this issue, offering clever and flavorful analogies to other corners of first amendment law on which more light has been shed.

Petitioners characterize the must-carry rules as posing more than an incidental burden on speech, likening the rules to the newspaper right-of-reply statute invalidated in Miami Herald Publishing Co. v. Tornillo, 418 U.S. 241, 94 S. Ct. 2831, 41 L. Ed. 2d 730 (1974), where the Supreme Court held that the enactment impermissibly interfered with the newspaper's constitutionally protected "editorial discretion." Toward this end, petitioners also offer the recent case of City of Los Angeles v. Preferred Communications, Inc., 476 U.S. 488, 106 S. Ct. 2034, 90 L. Ed. 2d 480 (1986), where the Court noted that the selection and organization of programs on cable television does involve some degree of editorial discretion. Id. 106 S.Ct. at 2037.2 See Brief for Joint Petitioners at 10-21.

The FCC counters by characterizing the must-carry rules as a commercial regulation that burdens speech in a far more attenuated fashion. Accordingly, the FCC argues, the must-carry rules are more appropriately analyzed under the standards set forth in United States v. O'Brien, 391 U.S. 367, 88 S. Ct. 1673, 20 L. Ed. 2d 672 (1968), where the Supreme Court stated that to be valid, a regulation incidentally burdening speech and not aimed at the suppression of free expression must advance a substantial governmental interest and must be no more restrictive than necessary to accomplish that end. O'Brien, 391 U.S. at 377, 88 S.Ct. at 1679. See Brief for FCC at 30-42.

The precise level of first amendment protection due a cable television operator is clearly an issue of much moment to the industry and ultimately to viewers. However, having closely analyzed the rationale for and workings of the new must-carry rules, we conclude that we again need not resolve this vexing question. Like the original must-carry regime invalidated in Quincy Cable TV, the new, scaled-back edition fails to satisfy even the less-demanding first amendment test of United States v. O'Brien whose use here is advocated by the FCC. See, e.g., Brief for FCC at 30 (describing O'Brien as "the correct test").3 We now proceed to offer our application of that test.

B. An O'Brien-Test Analysis of the New Regulations

In United States v. O'Brien, the Supreme Court stated:

[W]e think it clear that a government regulation is sufficiently justified if it is within the constitutional power of the Government; if it furthers an important or substantial governmental interest; if the governmental interest is unrelated to the suppression of free expression; and if the incidental restriction on alleged First Amendment freedoms is no greater than is essential to the furtherance of that interest.

391 U.S. at 377, 88 S. Ct. at 1679. Typically, analysis under United States v. O'Brien begins with an appraisal of whether the interest said to be served by a governmental measure is substantial. If it is, we proceed to the more delicate fact-bound issue of whether the means chosen are congruent with the desired end, or whether they are too broadly tailored to pass muster. See O'Brien, 391 U.S. at 377, 88 S.Ct. at 1679; see also Members of City Council of Los Angeles v. Taxpayers for Vincent, 466 U.S. 789, 804-05, 104 S. Ct. 2118, 2128-29, 80 L. Ed. 2d 772 (1984) (applying O'Brien test to uphold city ordinance prohibiting posting of signs on public property); Quincy Cable TV, 768 F.2d at 1454-62 (using the two-step O'Brien framework employed here).

In this endeavor we are mindful of the fact that it is a first amendment test we are applying. Although at times an O'Brien inquiry into an agency regulation may appear to resemble an exercise in administrative law analysis, the Supreme Court has often noted that the substantial deference due in the administrative context has little relevance when first amendment freedoms are even incidentally at stake. See, e.g., Members of City Council v. Taxpayers for Vincent, 466 U.S. at 803 n. 22, 104 S. Ct. at 2127 n. 22 (courts "may not simply assume that [an] ordinance will always advance the asserted state interests sufficiently to justify its abridgment of expressive activity"); Schad v. Borough of Mount Ephraim, 452 U.S. 61, 101 S. Ct. 2176, 68 L. Ed. 2d 671 (1981) (holding that no governmental justification existed to support application of a zoning ordinance to bar nude dancing); Heffron v. International Society for Krishna Consciousness, Inc., 452 U.S. 640, 658, 101 S. Ct. 2559, 2569, 69 L. Ed. 2d 298 (1981) (Brennan, J., concurring in part and dissenting in part) ("As our cases have long noted, once a governmental regulation is shown to impinge upon basic First Amendment rights, the burden falls on the government to show the validity of its asserted interest and the absence of less intrusive alternatives.").

We stress at the outset that both the justification offered by the FCC for its new regulations and the scope of those new initiatives differ rather markedly from the justification for and scope of the initial must-carry rules struck down in Quincy Cable TV.4 We therefore do not by any means accept petitioners' characterization, see Brief for Joint Petitioners at 1, of the new must-carry rules as mere imitations of those invalidated in Quincy Cable TV and thus deserving of a hasty execution. Although Quincy Cable TV supplies the structural framework for our analysis, the new must-carry rules are to be evaluated on their own terms: they should not suffer by dint of their association with the previous must-carry regime.

Our reservations about the new must-carry rules do, however, implicate both the substantiality of the governmental interest advanced and the narrowness of their design.

1. The Substantiality of the Governmental Interest

It may well be that upon a suitable record showing, the justification offered by the FCC, that interim regulations are needed to keep local broadcasts accessible to viewers while the new switch-and-antenna technology takes hold, would satisfy the O'Brien standard. See, e.g., FCC v. WNCN Listeners Guild, 450 U.S. 582, 594, 101 S. Ct. 1266, 1274, 67 L. Ed. 2d 521 (1981) (deeming "the policy of promoting the widest possible dissemination of information from diverse sources to be consistent with both the [Commission's] public interest standard and the First Amendment"); cf. FCC v. National Citizens Committee for Broadcasting, 436 U.S. 775, 795, 98 S. Ct. 2096, 2112, 56 L. Ed. 2d 697 (1978) (noting first amendment value of achieving " 'widest possible dissemination of information from diverse and antagonistic sources' ") (citations omitted). The difficulty is that here, as in Quincy Cable TV, the FCC's judgment that transitional rules are needed is predicated not upon substantial evidence but rather upon several highly dubious assertions of the FCC, from which we conclude that the need for a new saga of must-carry rules is more speculative than real. See, e.g., Home Box Office, Inc. v. FCC, 567 F.2d 9, 50 (D.C.Cir.), cert. denied, 434 U.S. 829, 98 S. Ct. 111, 54 L. Ed. 2d 89 (1977) (requiring agencies to present "a record that convincingly shows a problem to exist" in order to satisfy the "substantial interest" prong of the O'Brien test); see also Quincy Cable TV, 768 F.2d at 1455 n. 44 (noting Supreme Court cases requiring "more than an unsubstantiated assertion of the importance of the governmental interest"). Such speculative fears alone have never been held sufficient to justify trenching on first amendment liberties.

The agency's first questionable contention is that consumers are not now aware and cannot be expected to become aware in fewer than five years that the installation of an A/B switch could preserve their choice of programs:

[T]he perception [exists] that cable systems may be able to preclude access by their subscribers to off-the-air broadcast signals. This perception derives not from any inherent characteristic of cable service, but rather from cable subscribers' current expectation that broadcast signals will always be available as part of their basic cable service. This expectation is a direct result of the former must-carry rules, which, in fact, required cable systems to carry all available off-the-air broadcast television signals. The expectation that local broadcast signals will be carried by their cable system has caused many subscribers to perceive that there is no need to install or maintain the capability to receive broadcast signals off-the-air.

* * *

If we did not adopt interim must-carry rules now, until our long-term regulatory plan to educate consumers on the need for independent access to off-the-air signals and to make input selector switches available takes hold, harm to the public interest would ensue.

See Report & Order at paragraphs 121, 126.

The FCC, however, adduces scant evidence for its judgment of a widespread "misperception" among cable subscribers that the only means of access to off-the-air signals is through cable service. It puts forth no attitudinal surveys, or polls, suggesting the likely pace of consumer adaptation to the A/B switch technology. Nor does it offer analogies illustrating how swiftly consumers have incorporated previous electronic innovations. Such evidence might have shown what the FCC simply assumes here: that upon the disappearance of must-carry regulations, consumers would collectively fail to install with any dispatch the switches and antennas necessary to gain access to local broadcast stations, conceivably imperiling the survival of these stations and thereby depriving viewers of diverse broadcasting offerings.

The lone item of "hard" record evidence on which the FCC relies in support of its need for a five-year interim must-carry period is a study entitled, "Outdoor Antennas, Reception of Local Television Signals and Cable Television," prepared by the ELRA Group, Inc., for the National Association of Broadcasters ("NAB"), see J.A. at 205 (cited at Report & Order at p 124). NAB's members have long benefited from the existence of must-carry rules, and the organization, during rulemaking, strongly criticized alternative proposed regulations that it perceived as inadequately protective of broadcasters, such as reliance on the A/B switch. See Report & Order at p 47 (summarizing NAB's statement to FCC).

This NAB study is essentially a statistical compilation of survey results gleaned from a poll of 610 heads of cable households. The FCC cites this survey generally without pinpointing any specific parts of it that would strongly support the new must-carry regime, see Report and Order at p 124. We probe its findings in greater detail here, seeking to uncover and evaluate the particular material in it that may reinforce the agency's rationale.

Among its melange of disparate facts and findings, the study includes four items of information that arguably could be said to point to a need for interim must-carry rules: (1) only about 1% of cable subscribers presently have both the outdoor antenna and A/B switch needed to gain access to noncarried local programming in the absence of must-carry rules; (2) many cable viewers originally owning antennas have taken them down, because they were unsightly, and only about 10% of cable subscribers presently switch back and forth between cable and antennas; (3) a third of cable homes have video cassette recorders and thus may face some increased difficulty attaching the A/B switch; and finally, (4) about half of cable subscribers doubted that if local broadcast stations were dropped from cable they would buy what the survey termed a "special switch" enabling them to go back and forth between cable and their antennas. Primarily on the basis of these findings, and particularly the finding that relatively few homes are presently equipped with both antennas and switches, the report concludes that the transition to a world without must-carry could force consumers as a whole to expend millions of dollars. It does not, however, suggest that the new technology would be especially costly to consumers on an individual basis. Nor does it estimate how long it would take for most households to acquire and install the required switch and antenna.

Even accepting the NAB's findings as accurate, it requires an inferential leap of some distance to arrive at a need for five more years of must-carry. Only through the rosiest of broadcasters' lenses can the NAB study's first salient finding--that there is a dearth of antenna-and-switch setups in American households--be seen as pointing to the difficulty of installing such gear or to the inability of consumers to learn of their availability. More likely, the absence of such equipment from most homes reflects the obvious reality that, so long as the government requires cable companies to offer local broadcasting through the must-carry regime, such supplemental equipment is unnecessary. The FCC's own determination that the consumer misperception upon which it so heavily relies "is a direct result of the former must carry rules," see Report & Order at p 121, seriously undercuts the NAB's implication that the unavailability of switch-and-antenna gear is an endemic or long-term problem.

The NAB study's second finding, that few of those with switch-and-antenna capability currently switch back and forth between cable and broadcast with any regularity, can most reasonably be accounted for by the fact that, in a must-carry world, the need to do so is slight. Like the fact that few households have installed switches and antennas, this finding merely describes present reality without offering any glimpse into how the change of one key variable--the lapse of must-carry regulations--would affect that reality. As petitioner Leghorn, who appeared before the FCC during rulemaking proceedings, observes, see Brief for Petitioner Leghorn at 12: "Common sense suggests that consumers who want to receive an off-the-air channel will quickly observe that they may need to purchase (or reconnect) antennas should their cable system cease offering their favorite broadcast stations."The NAB study's third potentially relevant finding, that many cable subscribers own VCRs and thus would face somewhat complicated problems hooking up the switch-and-antenna, is readily dismissed as a grounds on which to justify the need for new must-carry regulations: the FCC itself, in its report explicating the new regulations, specifically discounts reliance on the VCR-interference theory. The Commission concluded:

We believe that any equipment compatibility problems can be overcome through relatively minor modifications to switching devices and that cable operators and other equipment suppliers can provide the information and/or assistance consumers need to install the switches for use with VCRs.

See Report & Order at p 167 (observing as well that "many of these concerns may become moot if television receivers begin to be manufactured with switching or interface devices built in"); see also Recon. Order at p 51 (noting that "[e]vidence that subscribers can make complex cable connections correctly is provided by the fact that there have been no widespread problems or difficulties encountered by consumers in installation of cable-ready VCRs and receivers ... many cable subscribers now are acquiring and successfully installing their own cable terminal/converter equipment").

The NAB study's final pertinent observation is that about half of the survey's respondents are unwilling to predict that they would ultimately purchase what the survey question termed a "special switch." Initially, we note that this characterization obscures somewhat the low price and easy installation of the A/B switch. Survey imperfections aside, however, this finding seems to us unpersuasive, for it almost certainly reflects merely the present consumer unfamiliarity with the switch and antenna mechanism. To the extent it does not, it may also reflect consumer disinterest in having access to off-the-air signals. Either way, this finding hardly explains why the five-year transitional period chosen by the FCC is necessary. The NAB's study thus provides only the spongiest of foundations for the FCC's asserted justification for its regulations.

In appraising the FCC's argument that the indelibility of consumer ignorance justifies the reimposition of must-carry rules, we are thus left to ask whether the FCC's contention is so obvious or commonsensical that it needs no empirical support to stand up. We conclude that it is not. For one thing, the FCC's own report elsewhere belies the agency's fears of viewer lethargy. The Commission notes:

There is evidence that video consumers are now becoming accustomed to switching between alternate program input sources. We observe that many cable systems now offer services through dual cables in order to provide greater channel capacity. Such systems employ switching devices to select between the two cables and often mark the switch positions with "A" and "B" designations. Cable subscribers apparently have accepted this switching arrangement and do not find it inconvenient.

See Report & Order at p 164 (emphasis added).

More generally, we simply cannot accept, without evidence to the contrary, the sluggish profile of the American consumer that the Commission's argument necessarily presupposes. In a culture in which even costly items like the video-cassette recorder, the cordless telephone, the compact disc-player and the home computer have spread like wildfire, it begs incredulity to simply assume that consumers are so unresponsive that within the span of five years they would not manage to purchase an inexpensive hardware-store switch upon learning that it could provide access to a considerable storehouse of new television stations and shows.5

Even were we to accept, however, the Commission's view that consumer ignorance cannot be readily eradicated, we have a second fundamental problem with the Commission's judgment that its interim must-carry rules are needed to advance a substantial governmental interest sufficient to support burdening cable operators' first amendment rights. The Commission relies heavily on its assumption that in the absence of must-carry rules, cable companies would drop local broadcasts. Experience belies that assertion. As cable operators reported to the Commission during rulemaking proceedings, see Report & Order at p 53, during the 16 months that elapsed between Quincy Cable TV and the reimposition of the modified must-carry rules, cable companies generally did not drop the local broadcast signals that they had been carrying prior to Quincy Cable TV.

The FCC responds that this constitutes "only limited direct evidence," and that in any event some cable companies did drop individual broadcast stations, see Report & Order at p 131. One might also speculate on behalf of the FCC that the inaction of cable companies after Quincy Cable TV may have partially resulted from their expectation that some new must-carry rules would inevitably emerge. Nevertheless, given Quincy Cable TV's vigorous denunciation of the breadth of the old must-carry rules, one can hardly assume that cable companies expected the FCC to reintroduce anything like the old sweeping must-carry requirements. Also undercutting the FCC's fearful assumption is the fact that both the Federal Trade Commission and the Department of Justice have concluded, in separate reports, that the absence of must-carry would not harm local broadcasting. See Report & Order at p 54 (noting Federal Trade Commission study, submitted in FCC rulemaking, that an analysis of 24 satellite television stations showed that "absent must-carry rules, cable systems can be expected to carry many or most local broadcast stations"); id. at p 55 (noting that Department of Justice also concludes that must-carry rules are not needed to foster localism); id. at p 114 (FCC acknowledges during post-Quincy Cable TV hiatus that "many cable systems are now providing locally originated programming services").

For these reasons, we conclude that the FCC has not demonstrated that the new must-carry rules further a substantial governmental interest, as the rules must to outweigh the incidental burden on first amendment interests conceded by all parties here. As we stated in Quincy Cable TV, "[a]t least in those instances in which both the existence of the problem and the beneficial effects of the agency's response to that problem are concededly susceptible of some empirical demonstration, the agency must do something more than merely posit the existence of the disease sought to be cured." 768 F.2d at 1455. The FCC error in this case was its failure to go that extra step here.

The second prong of the O'Brien test focuses on the congruence between the means chosen by the agency and the end it seeks to achieve. In this case, even were we convinced that the interest in whose name the FCC purports to act was more than a "fanciful threat," see Home Box Office, Inc. v. FCC, 567 F.2d 9, 50 (D.C.Cir.), cert. denied, 434 U.S. 829, 98 S. Ct. 111, 54 L. Ed. 2d 89 (1977), the new mustcarry regulations, because of their lengthy duration, are too broad to pass muster even under the O'Brien test.

If any interim period of must-carry rules is, in fact, necessary, the FCC adduces literally no evidence that this period must last for fully five years. Such a period is strikingly long in an industry that the FCC itself characterizes as "rapidly evolving." See Report & Order at p 133. In the absence of any empirical support for the new must-carry rules, the FCC falls back on what it terms a "sound predictive judgment," see Recon. Order at p 62, that it will take about five years for consumers to learn about the switch-and-antenna mechanism, and thus that a five-year transition period is needed during which the agency will provide consumer education.

We are, however, unpersuaded. In large part our reluctance to countenance reimposing must-carry rules for five years based on a "sound predictive judgment" that is never explained reflects our perceptions about consumer aptitude stated earlier. Such a guess about consumer instincts hardly presents the sort of issue where, "if complete factual support ... for the Commission's judgment or prediction is not possible," we should defer to the Commission's expert judgment. See FCC v. National Citizens Committee for Broadcasting, 436 U.S. at 814, 98 S.Ct. at 2121. It is wholly unclear to us why it should take five years to inform consumers that with the installation of a $7.50 switch and a television antenna they can view more local channels. The FCC report does nothing to shed light on this matter.

Additionally, we are skeptical--and the FCC's report says nothing to relieve this skepticism--that any consumer education campaign will have much impact so long as viewers can continue to rely on must-carry to get their fix of local broadcasts. It is entirely likely that not until the waning few months of the five-year must-carry regime would the FCC's admonitions about the need for switches and antennas begin to sink in, much as the existence of switches and antennas has largely gone unnoticed in a consumer population already accessed to local television as a result of must-carry in recent years. Opting for a five-year interim period therefore merely delays the inevitable, but almost certainly brief, period during which TV owners will learn of, purchase, and install the requisite equipment.6 We therefore find it difficult to defer blindly to the Commission's unproven belief that half a decade is necessary.

Our decision today is a narrow one. We hold simply that, in the absence of record evidence in support of its policy, the FCC's reimposition of must-carry rules on a five-year basis neither clearly furthers a substantial governmental interest nor is of brief enough duration to be considered narrowly tailored so as to satisfy the O'Brien test for incidental restrictions on speech. We do not suggest that must-carry rules are per se unconstitutional, and we certainly do not mean to intimate that the FCC may not regulate the cable industry so as to advance substantial governmental interests. But when trenching on first amendment interests, even incidentally, the government must be able to adduce either empirical support or at least sound reasoning on behalf of its measures. As in Quincy Cable TV, we reluctantly conclude that the FCC has not done so in this case, but instead has failed to " 'put itself in a position to know' " whether the problem that its regulations seek to solve " 'is a real or fanciful threat.' " Quincy Cable TV, 768 F.2d at 1457-59 (quoting Home Box Office, Inc. v. FCC, 567 F.2d 9, 50 (D.C.Cir.), cert. denied, 434 U.S. 829, 98 S. Ct. 111, 54 L. Ed. 2d 89 (1977)). Accordingly, we have no choice but to strike down this latest embodiment of must-carry.

So Ordered.