concurring.
Left to my own inclination, for the reasons set forth below I would deny both cross-motions for summary judgment and set this case for trial. My colleagues, however, are persuaded to opposite conclusions on the merits on the record presently before us. Were I to follow my own predisposition as a minority of one I would nevertheless prevail by stalemate, which should not be. Thus, obliged to choose between two forcefully argued but contradictory analyses of what I find to be a most ambiguous record — eloquent testimony, I believe, as to why a trial is necessary — in order that the case be decided I elect to concur with District Judge Sporkin: I am satisfied that, taken at face value, the evidence before Congress in 1992 of an impending demise of local broadcasting was at least “substantial” enough to warrant its prediction (a prediction, I note, to which we are instructed to accord a similar measure of deference whether we are convinced or not) in the absence of must-carry. Turner Broadcasting System, Inc. v. F.C.C., — U.S. -, -, 114 S.Ct. 2445, 2471, 129 L.Ed.2d 497 (1994) (“Turner”). I am also prepared to accept as minimal the burden must-carry imposes on the cable industry’s First Amendment rights of free speech in default of a genuine Fifth Amendment Takings Clause challenge based upon the fact that must-carry is for free.
In Turner, the Supreme Court vacated our judgment upholding the 1992 Cable Act,1 which we had also granted on cross-motions for summary judgment, and remanded the case to us for further proceedings consistent with its opinion. Finding genuine issues of material fact still to be resolved, — U.S. at -, 114 S.Ct. at 2472, the Supreme Court directed us to develop a more thorough factual record and to resolve any factual disputes remaining before passing upon the constitutional validity of the must-carry provisions of the 1992 Cable Act. Id. The ultimate issue, viz., whether the must-carry provisions of the 1992 Cable Act impermissibly infringe upon the First Amendment rights of free speech of cable TV operators and programmers, remains unchanged.
Specifically, we were directed to determine the factual validity of the Congressional premise upon which the Act was based, namely, that without the must-carry provisions, “the economic health of local broadcasting is in genuine jeopardy.” Id. at-, 114 S.Ct. at 2470. Then, if the evidence offered reasonable support for the premise, we were to consider the impact of must-carry on the cable operators and programmers affected. Id. at -, 114 S.Ct. at 2472.
Although both sides insist that the case is still susceptible of disposition on their cross-motions for summary judgment, the record they have assembled after extensive post-remand discovery belies it. They have been unable to agree on a common statement of the ease after several attempts to do so, and most of the data I can extract from their separate submissions, when cabined by the response from the other side, I find to be intractably ambiguous.
We have previously found, as did Congress, that by 1990 over 50 percent of American households received video signals via cable, having increased from less than half that figure a decade earlier, and the rate at which cable service is. expanding allows a prediction of still more, perhaps as much as 70 percent, by the turn of the century. Virtually all localities are served by a single cable operator of whom nearly a half are controlled by four or five conglomerates, known as “multiple systems operators” (“MSO’s”).
Defendants’ central premise, apparently accepted by Congress, is that the typical local cable operator is or will soon be one monopolist in the control of another, a component of an elite oligopoly of MSO’s who *753guard the electronic “gate” to most of America’s TV screens.
Of the market forces at work, we know that the lifeblood of broadcasting is advertising. Advertisers place their accounts with the medium offering the greatest dissemination. Cable companies compete with local broadcasters for advertising. It is therefore to the cable industry’s advantage to carry programming from which it will likely derive advertising revenue as well as subscription income. Interlocking ownership of cable operation and programming interests compounds the problem. Other factors being neutral, cable operators prefer to carry the programming of affiliated programmers in whose advertising revenue they share than that of competing broadcasters who pay them nothing.
How significant are these forces, operating in the economic universe premised by defendants and Congress? Here, the ambiguity of the truly uncontroverted facts I am able to quarry from the record precludes any definite conclusions in which I am sufficiently confident to submit to the Supreme Court as a basis for constitutional judgment.
For example, the defenders of the Act tell us that even in the first two post-Act years cable operators’ local advertising revenues increased by almost half and will more than double by the year 2001 to nearly $3.0 billion. Plaintiffs respond that virtually all of the cable industry’s income has been, is now, and always will be derived from subscriptions. Advertising accounted for less than five percent of the industry’s income in 1992 and is not likely to exceed 10 percent until well into the next century.
Be that as it may, defendants rejoin, in the five or so years antedating the Cable Act of 1992 over half of the nation’s local broadcast stations, particularly in the smaller markets, went from being profitable to otherwise. More than a dozen had ceased operations altogether, and nearly twice as many were in bankruptcy.
Profitable or not, the plaintiffs reply, in absolute numbers broadcasters increased from about 1200 to over 1500 in roughly the same period. Those in bankruptcy represent less than two percent of the total.
Plaintiffs reject the causal connection postulated by the defendants between cable carriage, or lack of it, and the prosperity vel non of local broadcasters as mostly conjectural, or no better than post hoc ergo propter hoc. There are many reasons why particular stations succeed or fail. As a general proposition, they say, local broadcasters with network affiliations particularly in the larger markets, were not only profitable but increasingly so from 1987 through 1992.
How important nevertheless do local broadcasters themselves regard must-carry to be? So important, defendants tell us, that more than a third of them immediately demanded it on one or more cable systems, foregoing the option to negotiate a fee for retransmission rights to their signals. Plaintiffs dismiss that assertion as standing the significance of the statistic on its head: the other two-thirds of all broadcast signals are now being retransmitted by cable operators who are paying the broadcasters for the right to do so, and some broadcasters who demanded must-carry on one cable system are being paid for retransmission rights on another. In any event, with or without must-carry, most cable operators have generally transmitted the signals of local broadcasters in their service areas as a matter of “viewer preference” because their subscribers want the service.
Turning to a consideration of the burden imposed by must-carry on the cable industry, how intrusive has it been on the cable operators’ autonomy to carry them programming of choice? We are told by defendants that somewhat less than 6000 cable channels nationwide have been conscripted by broadcasters invoking must-carry, whereas the cable industry’s total channel capacity has increased since 1992 by nearly 90,000 channels; for every channel pre-empted by a broadcaster 14 more have come on line. The cable industry responds that its technological advances have occurred unevenly across the country; more than two-thirds of the nation’s households today are served by cable operators who possess no excess channel capacity whatsoever.
*754Moreover, replete as this record is with equivocal data as it relates to competition between conventional broadcasting and conventional cable, it is also conspicuously barren of information about still other market forces that may have vastly more significance in the immediate future, and will bear directly upon such issues as how long must-carry should constitutionally be allowed to last. We are told little, for example, about the effects to be anticipated from direct broadcast satellite and telephone company competition, or impending legislation. Nor are we able to discern much about the attitudes of the public itself, assuming consumer preferences were of concern to Congress at the time of the 1992 Cable Act and remain so today.2
The Supreme Court has reminded us that in addition to our obligation to “accord substantial deference to the predictive judgments of Congress,” — U.S. at -, 114 S.Ct. at 2471, we do not have a “license to reweigh the evidence de novo, or to replace Congress’ factual predictions with our own.” Id. We are to ask only if Congress has drawn reasonable inferences based upon substantial evidence. I am also aware that neither side believes a trial is necessary. Nevertheless, the totality of the evidence of record here, with the myriad inferences that can be drawn from it, convinces me that a reasonable ■ trier-of-fact could return a verdict for either party, depending upon whose evidence best survived the adversarial process. In such circumstances, to my understanding, a trial becomes inevitable. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986).
Before WILLIAMS, Circuit Judge, and JACKSON and SPORKIN, District Judges. STEPHEN F. WILLIAMS, Circuit Judge, dissenting.Table of Contents
I.Introduction
A. Procedural History......................................................755
B. Overview...................■...........................................755
II. The Supreme. Court’s Mandate...............................................756
III. The Posited Interests.......................................................758
A. The Economic Health of Local Broadcasting...............................758
1. The continued viability of broadcast as a whole..........................759
(a) Total advertising revenue.........................................760
(b) Number of broadcast stations.....................................763
(c) Profits.........................................................766
2. Maximum possible expansion-of broadcast..............................767
3. The health of stations electing must-carry..............................768
■4. Survival of broadcast across all geographic markets .....................771
B. Unfair Competition.....................................................772
1. Anticompetitive behavior due to vertical integration.....................772
2. Anticompetitive behavior due to local monopoly regardless of vertical integration.......................................................776
(a) Independent programmers might outcompete broadcasters for access ..............................'.........................776
(b) Broadcasters as conduit competitors with cable......................778
IV. Narrow Tailoring...........................................................779
A. Must-carry’s Burden to Operators’ and Non-Broadcast Programmers’ Speech..............................................................780
B. Less Restrictive Alternatives............................'.................782
2. I observe that we do not know how many of those cable-wired households were able, and remain equipped, to receive high-quality over-the-air broadcast signals, nor, for that matter, how many are interested in retaining the option,
*7551. Leased, non-discriminatory access.....................................782
2. A/B or “input selector” switches ......................................785
3. The Century Rules..................................................788
V. Conclusion.................................................................789
I. Introduction
A. Procedural History
This case is on remand from the Supreme Court in Turner Broadcasting System v. FCC, — U.S. -, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994), with instructions to resolve a variety of factual issues that the Court held were pertinent to the plaintiffs’ contention that sections 4 and 5 of the Cable Television Consumer Protection and Competition Act of 1992 (“Cable Act” or “Act”), 47 U.S.C. §§ 534-35, violate their rights under the First Amendment. After extensive discovery yielding a record of tens of thousands of pages, the parties have cross-moved for summary judgment. For the reasons stated below, I would grant the plaintiffs’ motion and deny the defendants’.
The contested sections, together with the other provisions of the Act, permit a commercial television broadcast station to either (1) refuse to allow a cable system to carry its broadcast signal unless payment to the station is negotiated (“retransmission consent”), 47 U.S.C. § 325(b), or (2) force an area cable system to carry its signal even if the system does not want to (“must-carry”). 47 U.S.C. § 534. Non-commercial stations cannot elect retransmission consent, 47 U.S.C. § 325(b)(2)(A), but may demand carriage. 47 U.S.C. § 535. In any case of forced carriage, the station to be carried may demand one of several specific channel positions on the cable system in question. 47 U.S.C. § 534(b)(6) (commercial stations); 47 U.S.C. § 535(g)(5) (noncommercial stations).
We originally applied the “intermediate” standard of First Amendment scrutiny set forth in United States v. O’Brien, 391 U.S. 367, 88 S.Ct. 1673, 20 L.Ed.2d 672 (1968), and a majority of this court found the must-carry provisions valid under that standard. Turner Broadcasting System v. FCC, 819 F.Supp. 32 (D.D.C.1993). The Supreme Court affirmed that the must-carry rules were content-neutral and thus properly to be tested under O’Brien’s, standard, but remanded the case to this court for further factfinding because of a “paucity of evidence” necessary to discern whether must-carry met that standard. Turner, — U.S. at -, 114 S.Ct. at 2472. The Court noted that there was a lack of evidence “indicating that broadcast television is in jeopardy,” as well as a lack of “any findings concerning the actual effects of must-carry on the speech of cable operators and programmers” and “findings concerning the availability and efficacy of constitutionally acceptable less restrictive means of achieving the Government’s asserted interests.” Id. After extensive further development of the record it is clear that must-carry is not narrowly tailored to address any government interests that are actually at stake. The must-carry provisions burden substantially more speech than necessary to advance the government’s interests, and therefore cannot meet the standard expressed in O’Brien and in Turner’s specific mandate to this court.
B. Overview
This opinion works through a large number of claims and variations on claims, as well as detailed facts about the broadcast and cable industries. There is obviously a risk in such an enterprise that the trees will obscure the forest. Accordingly, I here give an aerial view of the forest, confining myself to my central conclusions.
1. The parties agree that there is no threat to the continued viability of broadcast television, either now in existence or looming on the horizon. Section III.A.l.
2. The defendants have offered several concepts by which to measure whether must-carry might advance some supposed governmental interest — such as maximizing the number and prosperity of broadcast stations. All but one of these (discussed in ¶ 3 immediately below) are circular, in that, once it is posited that must-carry would enhance the *756profits of the stations electing must-carry (a point that seems never to have been disputed), it is so obvious that must-carry will advance the interest in question that we do not believe the Court would have remanded for fact-finding if it had concluded that such interests were adequate to sustain must-carry. Sections III.A.2, III.A.3.
3. While the government has pointed to conclusory assertions that a significant number of stations have been materially injured by the absence of must-carry, (1) I believe that such evidence would be relevant only as evidence that there is some threat to the viability of broadcast television, a claim defendants emphatically do not make, and, in any event, (2) plaintiffs have met these conclusory assertions with substantial contradictory evidence. While I would not rule that summary judgment can necessarily be granted in favor of plaintiffs on the issue, it is absolutely clear, based on the detailed character of plaintiffs’ evidence, that summary judgment cannot be granted in favor of defendants. Section III.A.3.
4. The proposition that persons in cable households have no access to broadcast stations except as the cable operator may carry their signals is completely unfounded. Nothing about attachment to cable ipso facto cuts a person off from broadcast. Every cable household can secure as good access to over-the-air television as it had before adopting cable simply by (1) retaining or acquiring the sort of antennae that the uncabled 40% of households necessarily use, and (2) operating either an input selector switch on their “remote” or an “A/B” switch that can be acquired at a local hardware store for a trivial sum. Section IV.B.2.
5. To the extent that “fair competition” is threatened by cable operators’ characteristic monopoly over provision of cable services (and in some cases their vertical integration with cable program suppliers), the problem is readily cured by regulation under which all suppliers of programming not affiliated with an operator (i.e., independent cable programmers and broadcast stations alike) are given an entitlement to access at a price compensating the cable operator for the costs of carriage. Such systems are in place for industries with equivalent structural problems, such as gas, electricity, and wire telecommunications. Such a system involves far less intrusion on speech than must-carry while better serving the interests of fairness and signal diversity: As non-broadcast programmers competing for access over a cable system are not arbitrarily excluded by a fiat entitlement in favor of broadcast, but compete for access on an equal footing, the system enables consumers to choose between programs on the basis of the usual price/quality factors that they use to make other consumer choices. Section IV.B.l.
Accordingly, I now address the Court’s mandate to us to assess the reality of the harms to be cured and the fitness of the chosen remedy.
II. The Supreme Court’s Mandate
Congress’s overriding goal in enacting must-carry was to “preserve access to free television programming for the 40% of Americans without cable.” Turner, — U.S. at -, 114 S.Ct. at 2461. The goal is an important one, the Court said, because over-the-air television is a principal source of information and entertainment for a great part of the nation’s population. Id. at -, 114 S.Ct. at 2469. In fact, the goal touches upon a government purpose of the “highest order,” because “the widest possible dissemination of information from diverse and antagonistic sources is essential to the welfare of the public.” Id. at -, 114 S.Ct. at 2470 (internal quotations and citations omitted). This goal, as the Court saw it, might be jeopardized simply by the prosperity of the cable industry: to the extent that viewers were lured by cable away from broadcast (and to the extent that they abandoned the antennae and other paraphernalia necessary to watch TV over-the-air), the potential audience for over-the-air TV would shrink. Indeed, on these assumptions the potential viewership of any station refused carriage by all of the local cable systems serving the broadcast station’s area1 would, on average, shrink to *75740% of what it would have been. This in turn might shrink or at least seriously impair the broadcast industry, leaving non-cable viewers with materially reduced options. Thus, Congress was dealing with a special case of the general problem that competition from a new industry may reduce the economies of scale enjoyed by an established one, and thereby disadvantage consumers who use the goods or services of the established industry.
I also understand the Court to find the congressional action grounded in a related but distinct concern, focused more on the “unfairness” or anticompetitive characteristics of the cable industry’s competition with broadcast than on the deprivations that might be suffered by non-cable households. Id. The unfair or anticompetitive aspects of cable are said to flow from a combination of the cablecasters’ exploitation of their market power as the sole providers of cable service in their respective communities (as they are in the overwhelming proportion of cases2) and from their significant degree of affiliation with cable programmers, which create and market programs for release over cable. (Plaintiff Time Warner, for example, is a programmer, producing channels such as Home Box Office, and also a cable operator. Its pending merger with plaintiff Turner Broadcasting will broaden its tie to cable programming.) Here the contention is that cable companies might deny broadcast stations carriage on their systems simply to protect or enhance the advertising revenue the cablecasters enjoy from cable programming and to realize the cost advantages of using affiliates’ programming. To the extent that must-carry is addressed to this problem, as a kind of specialized antitrust regime designed for a possible anticompetitive injury, lesser impairments to viewer welfare might suffice to warrant government action — compared to the more fundamental threat to non-cable households’ access to over-the-air TV discussed above.
While these interests are important in the abstract, when the government seeks to legislate in the area of free speech “it must do more than simply posit the existence of the disease sought to be cured. It must demonstrate that the recited harms are real, not merely conjectural, and that the regulation will in fact alleviate these harms in a direct and material way.” Id. (internal quotations and citations omitted).
Thus, the Supreme Court found in Turner that pursuit of the goal of saving non-cable households’ access to free TV by means that impinge on free speech would be permissible only if that access were truly threatened — if “the economic health of local broadcasting is in genuine jeopardy” — and then only so long as the government’s remedy “does not burden substantially more speech than is necessary to further the government’s legitimate interests.” Id. (internal quotations and citations omitted). Similarly, the goal of correcting anti-competitive behavior can justify an infringement on free speech only if cable companies are in fact engaging in anti-competitive behavior (or if Congress could infer a serious risk of such behavior from circumstances such as industry structure) and then only if the scope of the proposed infringement — in this ease, the must-carry regime— reasonably fits the evil at hand.
This court’s first task in applying O’Brien scrutiny, then, is to ensure that Congress has drawn “reasonable inferences based on substantial evidence” in finding that there is a real threat to the survival of free broadcast television. (Section III.A.) In doing so we must accord substantial deference to the predictive judgments of Congress, making an independent judgment of the facts but not reweighing the evidence de novo. Id. at -, 114 S.Ct. at 2471. The court must also evaluate the presence or threat of anticompetitive behavior against broadcasting by the cable industry. (Section III.B.) Finally, it must consider whether must-carry is narrowly tailored to address the risk or risks — if any — found to be real. (Section IV.)
*758Defendants have asked us first to consider only the information before Congress as it passed the Cable Act. If that evidence were substantial enough to satisfy the defendants’ burden, then, they argue, we need look no further. Memorandum of the United States and the Federal Communications Commission in Support of Their Motion for Summary Judgment (“Government Brief’) at 4. Only if that evidence were insufficient might we look beyond. Id. at 5-6. I would reject this contention. The Supreme Court has frequently considered evidence outside the legislative record in constitutional review of statutes, see generally Kenneth Karst, Legislative Facts in Constitutional Litigation, 1960 Supreme Ct. Rev. 75, never, so far as appears, suggesting that the statute would be invulnerable to such evidence if it were supported by the legislative record — a record, one suspects, typically put together under the aegis of the statute’s proponents. In any event, defendants have deliberately chosen not even to present the record that was before Congress in the usual sense of the term. The materials they have assembled and labelled the “Congressional Record” are neither the Congressional Record as such nor the documents and testimony actually presented to Congress itself. Rather, they have included voluminous materials submitted to the Federal Communications Commission during the extended period that must-carry was before Congress. See, generally, Defendants’ Joint Statement of Evidence Before Congress. But if anything before the FCC was “before” Congress, it is hard to see in what sense anything in the world was not equally before Congress; all members of Congress have been continuously free to consider any fact extant in the world. As defendants have made no effort to isolate data that was peculiarly before Congress, they have made it difficult or impossible for us to apply the principle they invoke.
Defendants cite the Court’s statement that without “a more substantial elaboration ... of the predictive or historical evidence upon which Congress relied or the introduction of some additional evidence,” Turner, — U.S. at -, 114 S.Ct. at 2472 (emphasis added), it was impossible to ascertain whether the alleged threat to broadcast television was real enough, and read it as saying that must-carry must be upheld if either source satisfied the constitutional standard. I think that a strained reading, especially in light of the Court’s prior reference to the absence of any need for Congress to “make a record of the type that an administrative agency or court does to accommodate judicial review.” Id. at -, 114 S.Ct. at 2471. In any case, the data supplied by defendants, taken all together, fail to sustain the burden explicitly placed on them by the Court of “demonstrat[ing]” the reality of the harms and their “direct and material” alleviation by the remedy selected. Id. at -, 114 S.Ct. at 2470.
III. The Posited Interests
A. The Economic Health of Local Broadcasting
The parties strongly disagree over exactly what proposition concerning broadcasting’s health and viability this court was to test upon remand. The congressional finding at issue is whether, “absent mandatory carriage rules, the continued viability of local broadcast television would be ‘seriously jeopardized! ” Turner, — U.S. at -, 114 S.Ct. at 2470, citing § 2(a)(16) of the 1992 Cable Act, 106 Stat. 1460 (emphasis added). See also § 2(a)(10) (“A primary objective and benefit of our Nation’s system of regulation of television broadcasting is the local origination of programming. There is a substantial government interest in ensuring its continuation.”) (emphasis added); § 2(a)(12) (“There is a substantial government interest in promoting the continued availability of such free television programming, especially for viewers who are unable to afford other means of receiving programming.”) (emphasis added); Turner, — U.S. at -, 114 S.Ct. at 2461 (“the provisions are designed to guarantee the survival of a medium that has become a vital part of the Nation’s communication system, and to ensure that every individual with a television set can obtain access to free television programming”) (emphasis added).
At first glance these congressional assertions and Supreme Court reformulations of the government interest seem rather straightforward: Congress believed that the *759continued availability of free local broadcast television, a fixture of the American cultural landscape for over fifty years, was in serious jeopardy. The defendants have urged several other interpretations upon us, however, each entailing a contention that is far more modest — and therefore more readily proven. One is that the governmental interest identified as substantial in Turner is the interest in ensuring that persons without cable have access “to the greatest possible multiplicity of broadcast outlets.” Reply Memorandum of the United States and the Federal Communications Commission in Support of Their Motion for Summary Judgment (“Government Reply”) at 5 (emphasis added). That is, Congress intended broadcast to be as successful as possible. Such an interest is self-evidently in jeopardy without must-carry, since must-carry will financially benefit the stations that adopt it; their resulting gains will enhance the survivability of marginal stations and will enable them all to invest more in programming. Defendants also offer another analysis that is equally self-validating. In this variant they contend that Congress’s concerns about “continuation” and “survival” of local broadcast applied only to those stations electing must-carry; the health of those broadcast stations not invoking must-carry is claimed to be irrelevant to an evaluation of the health of local broadcast.3 A final possible interpretation holds that Congress might have intended to ensure that viewing areas across the country all enjoyed at least a substantial number of broadcast signals; it is small solace to a viewer in a location with few or no broadcast signals that viewers elsewhere have plenty. I evaluate each of these interpretations in turn, both for the likelihood that each is in fact what the Court embraced as an objective substantial enough to justify must-carry, and for the possibility that each is really, rather than merely conjecturally, at stake.
1. The continued viability of broadcast as a whole
Sections 2(a)(10), 2(a)(12), and 2(a)(16) of the Cable Act, quoted above, seem to assert on their face that the survival of local over-the-air broadcast programming is in question. These legislative findings emphasize the importance of a continuation of broadcasting and the fear that without must-carry “the economic viability of free local broadcast television and its ability to originate quality local programming will be seriously jeopardized.” § 2(a)(16).
In fact the defendants unequivocally concede, in the briefs and in the depositions of their experts, that the evidence they have introduced on remand does not show that the health of the broadcast industry as a whole is in danger.4 For example, the defendants’ distinguished chief expert, the economist *760Roger Noll, observed that he did not think that must-carry “in the current environment is properly characterized as something that has to do with the viability of the entire industry.” Deposition of Roger G. Noll (“Noll Deposition”) at 156. See also id. at 79 (“I do not believe that over-the-air broadcasting is going to disappear because of the absence of ‘must-carry’. If that’s what the sentence [“health of the industry”] is meant to imply, then indeed the health of the industry isn’t threatened.”); Deposition of Harry Shooshan, III at 79 (“So do you need must-carry to support the health of the broadcasting industry as [a] whole today, no.”); Deposition of Jack N. Goodman at 7 and 129 (agreeing that broadcasting was “fairly healthy overall” and as of 1989, “the TV industry as a whole looking at gross numbers was not in an economic decline”).
The parties offer several means of empirically assessing broadcast’s health, all tending to vindicate the above statements. First, since the mainstay of free commercial television’s diet is the sale of airtime to advertisers, changes in advertising revenue over time can reveal whether broadcast is thriving. Second, changes in the overall number of stations can be a rough proxy for whether the national broadcast system is in ascendance, plateau, or decline. If the number of stations has been holding steady or increasing, even in the period without must-carry, then it is difficult to see how the industry can be said to be in danger. Third, the government suggests that profits for a particular segment of the industry — UHF independents — are so abysmal as to put the health of at least that segment in substantial doubt. Although this latter data set relating to a special industry sector does not directly speak to the health of the industry as a whole, and does not appear to be offered as a bellwether for other segments’ futures, I treat it here to consider whether it calls for any important qualification of defendants’ concession about the robust health of the industry overall.
(a) Total advertising revenue
Plaintiffs submitted sets of data from 1981 to 1991 from which gross advertising revenues for broadcast, cable, and other industries can be analyzed, see Time Warner Entertainment Company, L.P.’s Corrected Appendix to Brief in Support of Its Motion for Summary Judgment Exhibit (“Time Warner Ex.”) 12, and defendants submitted the same data, together with the consumer price index for the relevant years, see Defendants’ Joint Submission of Additional Evidence in Support of Motion for Summary Judgment (“Defendants’ Additional Evidence”) Vol. VII.W Exhibits 668 & 672. I have derived from them a table of broadcast’s total advertising volume in constant dollars during that period, see App. # 1, and a graph, Graph # 1. In that ten-year period taken as a whole, broadcast’s overall gain in gross ad revenue was $3.7 billion in constant dollars.
*761[[Image here]]
After a rise in the early ’80s, gross revenues were more or less flat from 1986 to 1990, and fell off by about $2 billion from 1990 to 1991.5 In this pattern there are no extended declines, from which one could draw a reasonable inference that the broadcast industry was in jeopardy of losing its lifeblood. The only sharp decline is that of 1990-91, congruent with the 1990-91 recession beginning in the second half of 1990, cf. National Economic Trends, Federal Reserve Bank of St. Louis, Oct. 1995 at 4; early in 1991 an official of the INTV (the defendant trade association representing independent broadcasters) attributed the nascent drop to the recession and concluded that “the basic fundamentals of independent television are stronger than they have ever been.” Time Warner Ex. 5 (from TV Digest, Jan. 7, 1991).
The inference that broadcast is healthy, and not materially harmed by cable, is further enhanced when the focus is expanded to look at advertisers’ expenditures on broadcast’s competitors. The graph below, Graph # 2, shows expenditures on cable, broadcast, and broadcast’s two largest competitors throughout the period (newspapers and direct mail), and all other modes put together.
*762[[Image here]]
First, it is apparent that in 1988-91, the period of broadcast’s modest decline of $2.9 billion (in constant dollars), the newspaper industry, broadcast’s twin heavy in the advertising world, suffered a similar drop (actually $4.4 billion). And the total amount spent on advertising across all industries dropped by $7.6 billion. In 1990-91, the year of broadcast’s adversity, newspaper also suffered a distinct drop while direct mail and cable were only marginally higher. See App. #1.
Second, defendants have not even hinted at some theory as to how the woes of newspapers, direct mail and “other advertising venues” could be due to cable, and cable’s diminutive status makes any such claim implausible. For 1981-91 as a whole, broadcast gained $3.7 billion over 1981’s revenue to total $19 billion in 1991, while cable showed a gain of $1.4 billion, taking its 1981 revenue of $150 million up to total $1.55 billion in 1991. While cable’s percentage gain was sharp, it started from such a low base that its total impact was necessarily modest. And if one looks to 1990-91, the year of broadcast’s $2,000 million drop, there is only a $61 million gain for cable — obviously not a serious explanation.
Interestingly, trends in cable advertising suggest that must-carry has little influence on the competition between the two industries for ad revenue. One can usefully divide the period into two phases, 1981-85, during which the FCC’s must-carry regulations prevailed, and 1985-91, the period after they *763were invalidated.6 In the first period the cable industry’s annual ad growth averaged $149 million per year; in the second — when according to defendants cable was exploiting must-carry’s absence to leverage ad revenue from broadcast — cable’s average growth was lower in constant dollars, $137 million per year. This picture is confirmed when one puts the matter in terms of market share and thus abstracts away the effect of overall changes in advertising volume. Cable’s market share increased at the rate of 0.14% per year with must-carry (’81-85) and 0.13% per year without it (’86-’91). App. # 1. Alternatively, as the Commission’s so-called Century must-carry rules were in effect from June 10, 1987 to December 11, 1987, see Defendants’ Additional Evidence Vol. VII.N Ex. 338,7 one might divide the period into 1981-87 and 1988-91. Cable’s average annual gains in those periods were, respectively, $131 million and $158 million, the latter reflecting 1989’s unusually high increase of $241 million.
I note at this point the defendants’ argument that cable pulled its punches in the must-carry gap period, strategically refraining from drops or refusals to carry in order to defuse the political pressure for must-carry. Indeed, some officials in the cable industry trade association actually exhorted cable companies to refrain from changes. See, e.g., Defendants’ Additional Evidence Vol. VII.N. Ex. 321 (letter from NCTA president urging cable operators to minimize channel repositioning; “ ... a sensitivity to the political climate needs to be shown ... if we are to come through all this without unnecessary criticism.”); id. Ex. 330 (“By minding our social and political P’s and Q’s, we will be better equipped to deal with the touchy situation that will result with all of these deregulatory events happening at the same time.”); id. Ex. 333 (after D.C. Circuit struck down FCC’s must-carry rules in Century, NCTA president states that “[i]n the interim, and having in mind that this decision may well be appealed, we will advise our members generally to behave as if the carriage rules just struck down were still in effect.”); id. Ex. 344 (Viacom cable executive insists that local systems seek central approval before dropping a broadcast station “[s]o that we do not become an example of demonstrated abuse and thereby the justification for reregulation”).
But I find these statements an inadequate reason to distrust the record before us. First, defendants offer not a clue as to how the cable industry as an aggregate would overcome obvious collective-action problems. While the industry as a whole might benefit from such strategic self-restraint, one individual cable system out of more than 10,000 could hardly. expect that its snipping off a broadcast station or two would be likely to incur congressional wrath in the form of legislated must-carry. Second, to the extent that the constitutionally fatal aspect of must-carry is the absence of evidence of any visible threat to broadcast, then the rules would become constitutional the minute the industry was visibly threatened. Thus, under Turner, and assuming no other constitutional infirmities, the cable industry will have exactly the same incentive to hold back as it had in the 1985-91 period. Assuming the industry has the capacity to coordinate its self-restraint (as defendants must), it will have to do so into the indefinite future in order to avert enactment of a valid must-carry.8
(b) Number of broadcast stations
Another way to measure the health of the broadcast industry is through the number of stations. More stations presumably translate to more choices for viewers of over-the-air television. The Supreme Court found it significant that at the time it reviewed the case there was no evidence to show that “local broadcast stations have fallen into bankruptcy, turned in their broadcast licenses, curtailed their broadcast operations, or suffered a serious reduction in operating revenues as a result of their being dropped from, or otherwise disadvantaged by, cable *764systems.” Turner, — U.S. at -, 114 S.Ct. at 2472. The first two inquiries (bankruptcies, abandonments of licenses) address the total amount of broadcasting available throughout the nation, and an analysis of the number of broadcast stations over time suggests why the government on remand has not offered an overall count of the number of stations that have fallen into bankruptcy through noncarriage (relying instead on “evidence before Congress” pertaining to twenty-three independent stations, Plaintiffs’ Joint Response to Defendants’ Proposed Undisputed Facts (“Plaintiffs’ Fact Response”) ¶ 65). Indeed, defendants have conceded their inability to show that any station, commercial or public, turned in its license in the period without must-carry. See Defendants’ Response to Plaintiffs’ Joint Statement of Undisputed Facts Established by the Record (“Defendants’ Fact Response”) ¶¶ 57, 59. I have generated the following graphs, Graph #3 and Graph #4, depicting the change in the number of broadcast stations from 1981 to 1991 and more broadly from 1953 to 1994. They are based on data provided in the Declaration of Stanley M. Besen (“Besen Deck”), Ex. 5, which defendants have not contested in this respect.
[[Image here]]
These data show that the number of stations grew from 1,065 in 1981 to 1,532 in 1994, with a net growth in stations every year. In fact, the average net growth of 41 stations per year during must-carry’s absence from 1986-91 more than kept pace with the average growth of 39.4 stations per year from 1981-85, with must-carry in effect. If one takes 1981-87 vs. 1988-91 as the appropriate periods, the net annual additions of stations are 46 and 30, respectively. Although this suggests a lesser growth rate in the must-carry gap, annual growth of 2-3% in number of supplying firms is hardly anemic for a ma*765ture industry.9 Over the whole period 1981-91, the broadcast industry witnessed a 39% expansion of its total number of stations. Congress could not have reasonably inferred from this that the industry was in decline or even in jeopardy of any material decline.
Although the data on station increase comes from the Television and Cable Fact-book by way of plaintiffs, defendants do not dispute the increase. Rather, they turn to the weakest subset of stations — UHFs—and point to a decline in the growth rate in the period without must-carry. Graph # 4 presents the UHF stations broken out from total stations.
[[Image here]]
Besen Decl. Exs. 5-6. In the four-year period 1981-85, with must-carry, there was an average of 39 new UHF stations per year, as against an average of 35 new UHF stations per year in the five-year-period 1986-91, without must-carry. From this the defendants conclude that “this stunted growth of broadcast stations takes on greater significance when contrasted with the vast potential for broadcast television to expand.” Memorandum of the United States and the Federal Communications Commission in Support of Their Motion for Summary Judgment (“Government Brief’) at 36-37 (emphasis added).10 *766But the decline in growth rate is so modest, and the persisting actual growth so robust, that there can be no real claim that the numbers prefigure an eventual leveling off of the actual number of stations on the airwaves, much less a decline.
(c) Profits
The parties sharply dispute the relevance of different data sets that arguably relate to the prosperity of broadcast stations. The plaintiffs call cash flow margins the “predominant industry yardstick” and point to near-universal positive cash flow margin figures for stations both with and without must-carry. See Plaintiffs’ Fact Response ¶ 51. Defendants object that since many stations can have low (but still positive) absolute cash flows but high cash flow (percentage) margins, absolute cash flow data may be deceptive.11 See Defendants’ Fact Response ¶ 67.
The defendants, by contrast, direct us to median profits — a number that plaintiffs question, because, as they observe, it is subject to manipulation as station owners may deplete profits by payments to themselves in the name of salaries or perks. It is, moreover, a function of accounting conventions that may not capture essential characteristics. But one need not select among the parties’ respective analyses, because even the defendants’ favorite metric seems to afford them little support. They point to the dismal 1991 figures for UHF independent stations, which as a class are the least prosperous of TV stations. See Opposition of NAB and INTV to Plaintiffs’ Motions for Summary Judgment (“NAB Opposition”) at 70-74. They show pre-tax profits in dollars by percentile as follows:
Percentile
25% 50% (median) 75%
1991 (1,748,130) (401,090) 419,613
Defendants’ Additional Evidence VII.W, Ex. 669. In other words, the median station was showing a pre-tax loss of over $400,000 in 1991. Although the defendants do not offer similar figures for a succession of years, which might enable one to discern patterns with some clarity, they do supply comparable numbers for 1987, for half of which there was no must-carry and the other half must-carry under the “Century rules,” i.e., those in effect in the last half of 1987 and invalidated by the D.C. Circuit in late 1987. See Century Communications Corp. v. FCC, 835 F.2d 292 (D.C.Cir.1987), clarified, 837 F.2d 517 (D.C.Cir.1988). Putting both figures together one gets:
Percentile
25% 50% (median) 75%
1987 (2,542,540) (1,207,525) 74,802
1991 (1,748,130) (401,090) 419,613
Defendants’ Additional Evidence Vol. VII.W, Exs. 669, 670. In other words, in the period from the end of the Commission’s must-carry rules to the year before the onset of congressional must-carry, the profit picture of broadcast’s poorest relations improved markedly. Average pre-tax profit of the strongest 25% more than quintupled, losses of the median station fell by more than two thirds, ‘ and losses of the very weakest 25% fell by about a third. This improvement in the profit picture occurred, it should be recalled, in a period when the overall number of stations grew by over 8%, and the number of UHF stations grew by over 14%. One might expect per-station profit to fall in an era of such expansion — at least where, as was the case, aggregate advertising revenues were stable or declining. (See Graph # 1.) Indeed, the chairman of the INTV observed that part of the problem of a softened advertising market for these stations was that there are “too many hogs at the trough.” Time Warner Ex. 5 (from TV Digest, Jan. 7, 1991). See also id. (investment analyst observes that some *767individual independent stations may not have done as well because of the increase in the total number of stations); Defendants’ Additional Evidence Vol. VII.M, Ex. 288 at 38 (FCC explains that from 1986 to 1988 aggregate viewership of independent stations increased but revenues dropped, finding an explanation in “the increasing number of independent stations, which kept viewership per independent station low.”).
While the parties have offered no explanations of the low profitability of independent UHF stations as a general matter, the defendants’ offerings on profitability do nothing to sustain their burden of identifying the absence of must-carry as a threat to broadcast’s overall health.
If, then, the question squarely put before us by the Supreme Court — “whether the Government has adequately shown that the economic health of local broadcasting is in genuine jeopardy and in need of the protections afforded by must-carry,” Turner, — U.S. at -, 114 S.Ct. at 2470 — is, at it appears on its face, a question about whether the broadcast industry is in danger of failing or in peril of a downward spiral, the answer from both parties, and that indicated by the evidence, is No.
2. Maximum possible expansion of broadcast
The government claims that the Act’s retransmission consent provisions, 47 U.S.C. § 325(b), through which a broadcast station can opt to forego must-carry and instead withhold its carriage from a cable company until some form of payment is negotiated, “alone evidence Congress’s understanding that the broadcast industry as a whole would not perish absent the must-carry rules.” Government Reply Brief at 3. The government accordingly offers a different interpretation of promoting the “continued availability” of broadcast programming: that the substantial interest to be tested is whether, without must-carry, broadcast can expand as much as with it.
Thus, extrapolating from the Court’s remark, “Likewise, assuring that the public has access to a multiplicity of information sources is a governmental purpose of the highest order,” Turner, at -, 114 S.Ct. at 2470, defendants assert that must-carry can be upheld if appropriate to ensure that those without cable have access “to the greatest possible multiplicity of broadcast outlets.” Government Reply at 5 (emphasis added). I would reject this analysis. I do not read the Court’s statement as meaning that we are to find that lack of must-carry impairs the industry’s health merely on the basis of evidence that its presence would bring about a larger industry, measured by number of stations, or a more prosperous one, by any metric. If the Court meant that Congress had a substantial interest simply in increasing the prosperity of a significant number of TV stations (and thus, depending on FCC licensing policy, possibly increasing the overall number), it surely would have declared in Turner that Congress could reasonably infer such an effect; it would not have thought a remand necessary.
Nor do I accept defendants’ alternative formulation — that the Court meant us to examine whether must-carry would, at the margin, tend to produce a “higher quality” broadcast industry. The argument that must-carry would do so is simple. Increased potential audience implies an increased audience, and an increased audience implies both a greater ability to spend on programming (because of larger revenues) and a greater willingness to do so (because each dollar invested in program quality will, with a larger audience, yield a larger return). See Government Brief at 38-39; Declaration of Roger G. Noll (“NoU Declaration”) ¶36 (“[CJable carriage is positively associated with station revenues and with improvements in programming”). Again, the intuitive obviousness of these results precludes the possibility that the Court remanded the case to us to discover them.
Defendants alternatively invite us to look to the total number of stations allotted to broadcast by FCC fiat, arguing that Congress might have been concerned by a growth in the number of vacant channels; indeed, the number of vacant UHF channels increased from 157 to 198 between June 30, 1989 and June 30, 1991. Government Brief at 37 n. 81. But, as detailed above, the number of operating UHF stations increased substantially in that period, so that the claim *768based on vacant channels rests solely on the government’s own decisions to increase the number of theoretically available channels.12
Even if the spectrum were filled with broadcast stations, however, the government hypothesizes further possible growth of the broadcast industry, since total broadcast viewership could grow even if the number of stations remained constant. Here the baseline against which the government measures the need for must-carry is not (1) broadcast in the absence of anticompetitive behavior by cable, nor even (2) broadcast in the complete absence of the cable industry, but rather is (3) the maximum broadcast audience growth, including the component that cable has affirmatively made possible by enabling broadcast signals to be carried farther and more clearly than they could over the air:
The extensive cabling of the nation extended the reach of UHF stations because it overcame a substantial part of the “UHF handicap.” UHF stations had difficulty accessing all of their potential audience over the air because of the inferior signal propagation characteristics of the UHF band____ When viewers subscribed to cable, the UHF stations were received with greater clarity, and sometimes on a valued lower channel number. As a result, they reached a wider audience, inducing more stations to enter the industry____
Noll Declaration ¶ 40. See also Declaration of James N. Dertouzos (“Dertouzos Declaration”) ¶31 (local broadcasters are likely to benefit from an expanded audience due to enhanced signal reception and access to wider geographic markets). Thus, because cable enables its subscribers to receive “broadcast” signals of higher quality than they could otherwise obtain, the government argues that “cable carriage expands the reach of local broadcasters and contributes to the health of local broadcasting.” Corrected Memorandum of the United States and the Federal Communications Commission in Opposition to Plaintiffs’ Motions for Summary Judgment (“Government Opposition”) at 91 (emphasis added). The government employs parallel reasoning to defend the must-carry provision that forces many cable operators to carry UHF stations on a lower-numbered, and thus apparently more prominent, channel position than the UHF station’s own on-air position. See 47 U.S.C. § 534(b)(6); Defendants’ Joint Submission of Congressional Record in Support of Motion for Summary Judgment (“Defendants’ Record”) Vol. I.R., Ex. 85, CR 11503 (UHF station chafed at having to negotiate with a cable system for a valuable lower channel, absent must-carry’s forced positioning provision). These theories claim, then, that it is not enough for broadcast to be a net gainer from the presence of cable (though, to be sure, defendants offer no econometric evidence indicating the net effect); the harm to be cured is simply loss of the benefits that must-carry would afford when added to the other benefits afforded by cable in its absence.13 This “harm” of course is real by definition, and thus, again, cannot be what the Supreme Court intended for us to assess on remand.
3. The health of stations electing mustcany
Defendants appear to propose another concept of the interest to be advanced by must-carry, namely, the economic health of those stations that elect must-carry. They argue that the assessment of harm from must-carry’s absence must be confined to broad-*769east stations that would presumably be denied carriage or dropped absent must-carry since they currently rely upon it.14 For example, to counter plaintiffs’ evidence concerning gross broadcast advertising revenue across all stations from 1985 to 1991, defendants respond that these numbers include the revenue of network affiliates, which “are likely to be unaffected by the must-carry regime.” Defendants’ Fact Response ¶ 62. See also id. ¶ 73 (“[A] general statistic referring to the entire industry is irrelevant, given the marginal financial condition of UHF independents (the stations most often denied carriage absent must-carry).” Indeed, first the defendants reject statistics indicating the health of network affiliates because they “generally opted for retransmission consent, and the stations most at-risk absent must-carry are UHF independents and public stations,” id. ¶ 65. They then reject statistics regarding the health of UHF independents in general: “[I]t is not the health of the average independent station that is at issue here, it is the ability of stations that have been denied carriage or repositioned to provide the same quality and quantity of programming----”). Id. ¶66.
By narrowing the field of stations to the most vulnerable, the defendants attempt to identify stations whose financial health is truly hanging in the balance — those for which the added subsidy of free cable carriage could arguably make a critical difference. Again, the unremitting growth in the number of even UHF independent stations both with and without must-carry, see section III. A.l.(b) above, and the improved profit picture for them on defendants’ own data from 1987-91, see section III.Al.(c) above, belie the notion that significant numbers of stations are in jeopardy or that must-carry could greatly affect their fate on the margin one way or the other. All defendants have demonstrated is that, all else being equal, forced cable carriage is at least of some benefit to the broadcast stations so carried. But this alone can vindicate Congress’s interest only if we take the valid interest to be, circularly, maximum possible expansion for broadcast. See section III.A.2. above.
Defendants attach great importance to the Court’s observation in Turner that the government’s claim that must-carry was
necessary to protect the viability of broadcast television -rests on two propositions: (1) that unless; cable operators are compelled to carry broadcast stations, significant numbers of broadcast stations will be refused carriage on cable systems; and (2) that the broadcast stations denied carriage will either deteriorate to a substantial degree or fail altogether.
— U.S. at -, 114 S.Ct. at 2471.
A natural reading of this passage is that the Court reasoned that must-carry would be justified if its absence would cause the deterioration or failure of so many stations that one might infer a risk to the viability of broadcast television. Defendants reject any such reading, disdaining (as we have seen) any burden to show that broadcast television was in any jeopardy at all. Rather, they argue that evidence showing that the absence of must-carry materially injured a lot of stations would in itself show the requisite need for must-carry. Accordingly they point to statements “before” Congress (i.e., either made to a congressional committee or made to the FCC and deemed before Congress), in which individuals associated with the broadcast industry or with particular stations made statements asserting harm to a station from the absence of must-carry. See, generally, Defendants’ Joint Statement of Evidence Before Congress.
Plaintiffs have sought to respond as to each station identified by defendants in briefs or evidence as having suffered — 142 commercial television stations (out of a total of 1138 in 1992) and 141 public stations (out of 368). See Appendices to Time Warner *770Reply Memorandum in Support of Motion for Summary Judgment. The number 283 is surely not insignificant, so that, if (1) defendants are right that the Court used “significant” in a sense completely disembodied from the “viability” of broadcast, and (2) the assertions “before Congress” were valid, defendants would prevail on this point. I have already explained why I believe the first necessary condition is not satisfied; I now comment briefly on the implausibility of the second.
In its Broadcast Station Rebuttal Time Warner meticulously recites evidence relating to the nearly 300 stations that at one point or another defendants claimed were victims of the absence of must-carry. For example, one station said to have “gone dark” because of being denied carriage appears actually to have gone dark for the very understandable reason that its transmitter was hit by a tornado. See Time Warner Broadcast Station Rebuttal at 134-35. Moreover, despite the station’s travails with cable, its owners found a buyer at the price of $1.5 million (licenses, of course, are issued free) in May 1990 — the depths of must-carry’s absence. Id.
If defendants’ station-by-station rebuttal is accurate, see Time Warner Reply Memorandum in Support of Motion for Summary Judgment, Appendices A & B, for a summary, scarcely any of the 283 alleged victims has been seriously jeopardized by the absence of must-carry. Defendants make some attempt at a station-by-station refutation, see Response of the United States and the Federal Communications Commission • to Time Warner Entertainment Company, L.P.’s Broadcast Station Rebuttal, while continuing to point to the assertions “before Congress.” See Government Reply at 12 n. 17. I have not examined the station-by-station data in enough detail to say whether, under the standards of Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 247-52, 106 S.Ct. 2505, 2509-12, 91 L.Ed.2d 202 (1986), summary judgment in favor of plaintiffs would be proper on this issue. I can, however, unequivocally say that plaintiffs have at a minimum put the individualized victimization accounts into material dispute, so that if defendants’ reading of the Court’s “significant numbers” passage is correct, a trial would be necessary to apply that reading.
In considering whether the data on drops and carriage refusals tend to demonstrate a threat to “the continued viability of local broadcast television,” Turner, — U.S. at -, 114 S.Ct. at 2470, I note that the defendants repeatedly characterize the stations under threat of drop — i.e., those likely to elect must-carry — as marginal. See, e.g., Noll Deposition at 17-18; id. at 60 (absence of must-carry would lead to dropping of stations of “relatively low audience rating”); id. at 221-22 (the less popular a station the more important is must-carry to its welfare); id. at 307-08 (nothing in his declaration or testimony says cable systems will drop the “more popular or popular over-the-air televisions stations”); Government Reply at 10 (“these same stations are particularly weak financially”); Government Brief at 35 (discussing “marginal financial position of these stations”). Indeed, they stress their experts’ observations that the audience appeal of the stations’ programs (whether measured in number of listeners or intensity of preference) will not be enough to seriously enhance a cable operator’s subscriber revenue (via more subscribers or higher fees). See, e.g., Noll Deposition at 60-63; id. at 142 (drop of third noncommercial outlet or weak UHF station will not affect decisions to subscribe); id. at 166 (similar); Second Declaration of Tom Meek ¶ 34. Plaintiffs have offered evidence that the average cable subscriber was served by a cable system carrying local broadcast stations accounting for about 97% of television ratings in non-cable households. Besen Deck at 41; Plaintiffs’ Joint Statement of Undisputed Facts Established by the Record (“Plaintiffs’ Facts”) ¶ 9. While defendants bristle at this statistic, and suggest that there are defects in the type of Nielson data that support it, Defendants’ Fact Response ¶¶7-9, they offer no alternative figure. And their own contentions — as to the marginality of electing stations and as to the improbability that they would beef up subscriber revenues — implicitly recognize that what is at stake without must-carry is a diminutive share of total television viewing.
*771I also note that defendants had claimed that without must-carry cable operators often substituted even more poorly-rated cable networks for broadcast stations that admittedly were already-poorly-rated. Their first attempt at demonstrating such a trend (presumably to show an anticompetitive motive in such a substitution) neglected to account for the fact, all else being equal, a cable network carried by a given system will have lower ratings than a parallel broadcast station on that system because the broadcast station is additionally available to the 40% of viewers in the area who do not subscribe to cable. See Plaintiffs’ Facts ¶ 176,177. In endeavoring to correct for this, defendants’ expert Tom Meek tried to compare (as we read his declaration) the ratings on cable of the least popular cable programs with the ratings on cable of broadcast stations added pursuant to must-carry, and found that the must-carry stations had ratings higher than many of the low-ranked cable programs. See Reply Decl. of Tom Meek at 11-12. But this comparison ignored the issue of what fraction of cable systems carried the cable programs being compared with must-carry broadcasting. Government counsel acknowledged at oral argument that the comparison still suffered apples/oranges problems, saying in defense that they had used the data they had. See Tr. 167; Defendants’ Fact Response ¶ 177. Defendants now assert only that there is no evidence that the must-carry programs “are less popular” than the marginal cable programs that they replace. Id.
Ultimately the parties seem to agree that, for both marginal cable and broadcast stations, operators’ decisions on what to air will often not rest on either’s meager ratings. The First Amendment right at stake is the operators’ freedom to air what they choose— broadcast or cable — having to answer only to their subscribers, not the government. The risk that cable operators may exercise that right for anticompetitive purposes may well permit government intervention, see section III.B below, but in the absence of a threat to broadcast’s general viability there is no other ground permitting government veto power over a cable operator’s selection of services.
4. Survival of broadcast across all geographic markets
A congressional interest in preserving access to free television programming for the 40% of Americans without cable, see Turner, — U.S. at -, 114 S.Ct. at 2461, might be thought in jeopardy if areas with only a handful of stations were threatened with the loss of just one. In such a situation, the harm to a diversity of viewpoints — at least for the people in those areas — would be far graver, and therefore worthy of remedial intervention, than if carriage drops and refusals, and their adverse effects, were concentrated in busy television markets.15 Or, to look at the other side of the coin, it would be hard to call the local broadcast unhealthy, or to find its health at risk, if one found that all localities were enjoying the benefits of a significant number of stations.
As it turns out, though, if there is any pattern to carriage refusal, it is concentrated in the most crowded broadcast markets. See Defendants’ Fact Response ¶ 11. For example, the government cites an FTC study which found that the proportion of local broadcast stations carried fell as the number of available local signals increased. Government Opposition at 15. The study found that when three local signals were available, all were carried; when nine local stations were available, about six were carried. Id. Thus, those areas with the fewest broadcast stations have the greatest proportion of voluntary carriage and the least need for forced carriage. It is only those areas in which the airwaves are relatively busy — areas in which concern about the health of local broadcasting would be at a minimum — that cable carriage becomes an issue. Where stations are few and the loss of only one theoretically significant, the defendants’ figures indicate that cable carries the signals voluntarily. Further, from 1981 to 1992 not one local *772television market out of the 209 geographic ADIs experienced a decline in the number of broadcast stations available. Besen Decl. Ex. 11. From 1987-92, the years of falling advertising revenues and the absence of must-carry, the Boston ADI dropped from 15 to 13 stations (it had 9 in 1981) and four other ADIs lost a single station. The other 204 ADIs had either the same or more stations. Id. The continued overall broadcast station growth during what has been claimed to be the industry’s darkest hour, then, cannot be said to conceal failures in certain geographic areas that are offset by growth in others.
Summary
I conclude that there is insufficient evidence from which Congress could conclude that “broadcast television is in jeopardy,” Turner, — U.S. at -, 114 S.Ct. at 2472, under any interpretation of that phrase consistent with the Court’s analysis. Only if the government’s interest is taken to be the maximum growth of broadcast, or enhancement of precisely the stations that would elect must-carry against the will of cable operators — interpretations of Turner’s mandate that are manifestly implausible — can the posited harms be viewed as real rather than conjectural.
B. Unfair Competition
The defendants assert two related bases for inferring that cable operators will make “unfair” or anticompetitive carriage decisions at the expense of broadcasters. Both depend on the characteristic monopoly of cable franchisees over provision of cable in their areas. The first theory looks at the combination of that market power with the vertical integration between operators and cable programmers, and argues that vertical integration will lead operators to unduly favor their affiliates’ programming over that of broadcasters. The second theory relies solely on the typical cable operator’s monopoly over cable, and argues that various features — ranging from cable operator hopes of predation to artificial distortions in the treatment of advertising — will thwart competition and deny broadcast consumers the access to broadcast that they would enjoy in a perfectly competitive market. I here address these claims one by one. I conclude that — at least without trial — it is impossible to preclude the possibility of a risk of anticompetitive behavior, the correction of which would constitute a substantial interest. This would not prevent me from granting summary judgment for plaintiffs, however, for I find in section IV that must-carry is plainly not a remedy narrowly tailored to any such risk.
1. Anticompetitive behavior due to vertical integration
Defendants argue that a vertically integrated firm controlling a monopoly bottleneck may not make efficient decisions about what to let through the bottleneck, unduly favoring the products of its own affiliates. Thus, they say, unaffiliated suppliers (here creators of programming, including broadcasters) will be disadvantaged in trying to reach the market on the other side of the bottleneck. For purposes of this analysis I accept not only the undisputed point that cable operators characteristically have a monopoly over provision of cable services in their areas, but also the more questionable proposition that cable operators have a monopoly over the provision of video signals to households subscribing to cable. As a person who subscribes to cable does not, in fact, ipso facto lose access to over-the-air television, but can continue to reach it merely by operating a so-called “A/B” (or input selector) switch and by maintaining whatever antenna may be necessary (see section IV.B.2 below), the latter proposition would require additional evidence. I nevertheless indulge it in favor of defendants for these purposes.
Defendants’ bias argument runs as follows. First, a significant number of cable operators are affiliated with programming networks. See, e.g., Government Reply at 30-31 (saying that as of 1994 operators serving more than 70% of cable subscribers held vertical interests in cable programmers); Plaintiffs’ Fact Response ¶ 88; see also id. ¶¶ 25, 26 (indicating that cable operators have ownership interests in 39 of 68 nationally delivered cable video networks). Second, to a 100% vertically integrated operator/programmer the real marginal cost of using “its own” program is *773close to zero: it costs almost nothing to make another copy (any stated charge is a wash on the firm’s collective books), and use by the affiliated cable operator does not in any way limit the program-producing branch’s opportunities to sell identical copies of the same program to other operators. The cost of purchasing similar programming from an independent network would be positive, even though the independent network can also make extra copies of its already-produced programming at near-zero cost. All else being equal, this appears to tilt the balance arbitrarily in favor of the affiliate’s programming. For example, as between programming produced by an affiliate that would boost subscriber revenue by 100 and that produced by a non-affiliate that would boost revenue by 120 but was priced at 30, the cable operator might well choose the affiliate’s programming over that of the unaffiliated competitor,16 with a possible loss to consumer welfare that is not offset by any cost-saving anywhere else in the economy. See Noll Decl. ¶ 17; cf. Dertouzos Deck ¶ 6.
As just mentioned, the competing independent programmer’s marginal cost of allowing distribution by any one cable operator appears at first blush also to be zero. But a programming firm selling all its products at marginal cost would not recover its total costs. To stay in business the independent must secure some sort of positive price, so it will necessarily charge such a price. That accounts for the hypothesized price of 30 in the example above. The careful reader, at this point, may ask why the unaffiliated supplier does not simply lower its price to this firm to somewhere between 20 and zero. At any such price both parties will gain. The unaffiliated supplier will make more money than without the sale, and, since its product contributes 20 more to revenue than does the inside product, the vertically integrated cable operator will also improve its position. Defendants do not explicitly address this question, but I will try to supply a response. Clearly the transactions costs of the hypothetical bargain may often be high. The outside supplier will typically not have reliable information on the extra value that its product contributes over and above the integrated firm’s inside product, so the relationship invites bluffing that may obstruct a bargain. Moreover, the outside firm will reasonably fear getting a reputation for undue willingness to cut its prices; if it keeps getting chiseled down to the environs of marginal cost (i.e., zero), its hopes of being a profitable enterprise will fade.17
Before complicating the example with advertising (an additional feature of defendants’ analysis), one should reexamine this account. Because by hypothesis the marginal cost of producing all copies after the first one is about zero for any producer (regardless of integration), one knows that the program-producing industry will not survive at all on a stand-alone basis unless firms sell above marginal cost. Yet, as soon as one recognizes this, one is saying that the conditions for efficiency, in a first-best sense, cannot possibly be satisfied. One should, therefore, be extremely cautious about concluding that the complications added by vertical integration will necessarily make matters worse. See R.G. Lipsey & R. Kelvin Lancaster, The General Theory of the Second Best, 24 Rev. Econ.Stud. 11,11 (1956); cf. 3 Phillip Areeda & Donald Turner, Antitrust Law ¶ 725e (1978) (vertical integration of successive unregulated monopolies likely to improve welfare).
Defendants press a particular version of the integration-bias argument that focuses on cable operators’ quest for advertising revenues. This, of course, is the setting of most interest to broadcasters, because, by tradition at least, they gain their revenue from ads, not from sale of programming.18 For *774example, in some cases programmers sell programs that include ads; the programmers, not the operators, are paid for the ads by advertisers on the basis of expected viewership. (In fact networks and operators make varied arrangements for allocating the returns from advertising.) See Noll Decl. ¶¶ 15-17, 23, 24. A fully integrated system will reap the revenues for ads sold by its programming division if its conduit division runs the program (and thus will count them as a plus factor in selecting programming), whereas the system would not reap equivalent revenues received by an unaffiliated program supplier for ads contained in that supplier’s programs. Id.
In a case where the affiliate’s program would yield the network 100 in subscription revenue and the non-affiliate’s 120 (i.e., the same numbers as in the earlier example), and the outside alternative is offered at 30, and in which the affiliate’s program would generate 20 in ad revenue and the non-affiliate’s 30, the system is likely to choose the affiliate’s program because this allows it both to acquire the program at no cost and to capture the ad revenue associated with the program. But by hypothesis the affiliate’s program still has less viewer appeal (as measured by subscribers’ willingness to pay and to submit to advertising).
This second example appears to add nothing material to the earlier one. Ad revenue is largely a function of viewership (see, e.g., Noll Decl. ¶ 36), itself a function of program popularity, and is thus simply one way of realizing the return on the creation of programs. In this latest example the outsider’s program is more attractive than the inside alternative (by 20), yet the firm may prefer the inside product. The main difference is that both products, besides adding 100 and 120 to subscriber revenues, are so appealing that they will do so even though the viewers must submit to substantial advertising.
But the ad revenue’s apparent enhancement of the inside product’s advantage is an illusion. The outsider, in the hypothetical, has simply withheld a portion of the product’s value, just as would an auto dealer who offered a car for sale without wheels. As moderately alert customers of the auto dealer would take that into account in agreeing on a price, so would our integrated firm. The range of possible bargains between the parties is essentially unchanged. Thus, if the outsider agrees to allow the integrated firm to reap the ad revenue, both of the parties can (as before) improve their positions by any sale in the range of zero to 30. (E.g., at 10, the integrated firm gets 120 in subscriber revenue and 20 in ad revenue, minus 10 paid the outside supplier, thus doing 10 better than if it used its own programming.) If the outsider insists on keeping the ad revenue, i.e., is already 30 ahead, then the potential gain to the integrated programmer/operator from buying outside drops by 30, so that it must be paid something between 30 and zero.19
There is considerable question whether the alleged “bias,” once analyzed, is actually a social problem at all. The orthodox thinking on vertical integration by an unregulated monopoly is that such integration normally adds nothing material to the distortions implicit in the monopoly itself. A monopoly is able to achieve a single monopoly profit on its sales, and its ownership of resources supplying an input normally has no bearing on the extent to which the price of its final product will exceed the competitive price. See 3 Areeda & Turner, supra., ¶ 725. Indeed, vertical integration between a monopolist and an imperfectly competitive upstream industry may reduce the economic distortions attributable *775to the monopoly, as the integrated firm will base its output decisions on true calculations of the marginal cost of its inputs, whereas a separate monopolist would base its decisions on cost figures inflated by the independent supplier’s selling at prices reflecting its need to earn a profit on a stand-alone basis. Id. ¶ 725e. Here, however, in view of the special importance Congress attached to increasing the range of programming suppliers able to reach the viewers, see, e.g., section 2(a)(6) of the 1992 Cable Act (“There is a substantial governmental and First Amendment interest in promoting a diversity of views provided through multiple technology media”), the evidence might support a finding that the alleged “bias” constituted a market distortion in whose possible correction there might be a “substantial governmental interest.”
The plaintiffs contend that any such bias of cable operators in favor of affiliates is entirely theoretical and that the evidence shows that cable operators have in fact displayed no bias against broadcasters. They point to the finding of defense expert Dertouzos that one study supplied no statistically significant evidence that a vertically integrated operator is less likely to carry a broadcast station than an unintegrated one. See Defendants’ Fact Response ¶ 102; Dertouzos Rebuttal Deck ¶¶ 1(a), 3-11. The defendants, however, point to instances suggesting that the bias is at work, such as the overwhelming adoption of a TCI network by TCI systems, as opposed to radically fewer adoptions by non-affiliates. Government Opposition at 21; see also Noll Deck ¶¶ 18-19 (describing studies finding that vertically-integrated cable systems tend to favor their own affiliates in deciding which channels to carry).
Moreover, because of the background fact of government regulation of cable’s rates, vertical integration poses a clear risk to a substantial interest. A vertically integrated monopolist subject to price regulation may be able to disguise its true costs by slipping ones that in reality belong to the upstream market into its apparent costs in the regulated monopoly sphere, thereby inducing the regulators to raise its rates. Since regulation is presumably aimed at holding the monopolist’s prices below the profit-maximizing level, this price increase nets the firm additional profits (i.e., the net gain per sale will not be fully offset by lost sales). See 3 Areeda & Turner, supra, ¶ 726; Roger Blair & David Kaserman, Antitrust Economies 290 (1985). Further, on a widely held view, the firm is likely to use these profits to undercut unaffiliated upstream suppliers. See Daniel F. Spulber, Deregulating Telecommunications, 12 Yale J.Reg. 25, 60 (1995).20 To put this into the cable context, the argument would be that vertically integrated cable companies would gain an unjustifiable competitive edge on unaffiliated programmers (including broadcasters), to the detriment of viewers’ ability to choose quality programming.21
Thus I conclude that Congress could reasonably believe cablecasters’ vertical integration with programmers constituted a problem — especially if one takes into account the special congressional concern for “promoting a diversity of views” and the price-regulated status of cablecasters — for which it might seek a fix.
This finding in favor of defendants should also apply to must-carry for public television stations. Here the analysis is slightly different, because public stations do not rely on “ads” in the conventional sense of the term. Apart from their government-provided subsidies and foundation grants, they depend for revenue on viewer contributions. But, as is true for commercial television and its ad *776revenue, the yield in contributions is necessarily dependent on the number of viewers and the intensity of their preference for the programs. There is no reason to doubt that, insofar as there is a risk that cable operators may disadvantage unaffiliated programmers for the reasons discussed above, parallel reasons apply for public television, to the possible detriment of public stations and their viewers.
2. Anticompetitive behavior due to local monopoly regardless of vertical integration
Defendants also argue that even non-vertieally-integrated cable operators will unfairly discriminate against broadcasters. Such discrimination might happen in two ways, the first grounded in broadcast’s potential as a programming provider to cable, the other in broadcast’s competition with cable as an alternate video conduit. First, because cable programmers can bring “ad avails” — blank spots within a program into which the cable company can insert ads — to the carriage negotiating table while broadcasters cannot, cable operators will, as defendants see it, improperly favor the cable programmers. See Noll Deck ¶24. Second, broadcast stations and cable operators are in competition. By denying carriage to a broadcast station, an operator can limit the station’s audience, thereby reducing its potential to earn ad revenue. To the extent that ad revenue dries up, the quality of the station’s programming may suffer, and some viewers may switch to cable. At the limit the station will be driven out of business, again to the potential profit of the cable operator. I address these issues in turn.
(a) Independent programmers might out-compete broadcasters for access
There are a variety of ways in which the manager of a television conduit — broadcaster or cable operator — might choose what programming to air. Some programs or complete channel feeds will be so popular that conduits will be willing to pay for them. Both cable operators and broadcasters do so frequently, and the parties agree that programming purchase costs turn out to be a large part of the typical broadcaster’s budget. Plaintiffs’ Fact Response ¶ 68. Since broadcasters are programmers, they are also in the business of selling programming. Broadcasters nowadays do so regularly— whenever a broadcast station demands retransmission consent before permitting a cable operator to air its programming. The conduits expect to cover the costs of program acquisition by various means. Cable operators may expect to cover them through enhanced subscriber revenue, either by an increase in the number of subscribers or an increase in per subscriber fees or both. They also may use the programming value to garner ad revenues; part of the transaction would be for the programmer to offer the operators blank spaces in which to insert their own ads — spaces known in the trade as “ad avails.” As broadcasters have no ability to prevent viewers from tuning in (apart from scrambling techniques), they of course get their return solely in the form of enhanced ad revenues.
Other programming is not so directly popular. In such cases the programmer might offer to pay the conduit instead of the other way around. In transactions of this sort the conduit makes its money from these very payments, typically (perhaps always) foregoing the opportunity to earn money by additional ad revenues. (By hypothesis these programs are ones unlikely to generate extra subscribers or higher fees per subscriber.) The record contains several undisputed examples of broadcast station conduits that earn their money through such alternatives to traditional advertising sales. Some stations, for example, are paid lump sums or commissions to carry home-shopping feeds, or are paid by religious organizations to carry religious programming. See Time Warner Broadcast Station Rebuttal at 33 (KMCI paid by the hour for programming blocks by Home Shopping Network); id. at 60 (KTFO leased its daytime operations to Tulsa Christian Television); id. at 78 (LeSea Broadcasting Stations — similar); id. at 85-86 (similar; purchaser of all of WRDG’s air time, “Jack Rehburg Ministries,” found its own sources of funding, i.e. donations, dried up under some combination of lost cable carriage and the onset of televangelist scandals).
*777Both sides concede that the broadcast stations most likely to avail themselves of must-carry have comparatively low ratings. Such stations are unlikely to be sought after by cable operators, especially because on the margin they cannot be easily expected to affect cable subscribership or rates. See NoU Deel. ¶ 19; Noll Rebuttal Deck ¶20. Comparatively low-rated cable programming might have an edge over a broadcast station because the programmers offer the operators ad avails. Defendants argue that these ad avails may be especially appealing since Congress has regulated the rates cable operators may charge their subscribers. Thus ad revenue has a possible advantage — both in availability and in freedom from regulatory hassle — not shared by higher subscription fees.22 But broadcasters cannot offer ad avails at all. This is true, so far as the record shows, solely by the terms of the commercial must-carry rules themselves; 47 U.S.C. § 534(b)(3) requires cable systems to run unaltered copies of the broadcast signals they carry. See Memorandum of National Cable Television Association Inc. in Opposition to Defendants’ Motion for Summary Judgment at 17. Therefore, all else being equal, the possibility of added opportunities to sell ad time gives low-rated programming offered by independent programmers an appeal to operators that low-rated broadcast programming cannot match.
The parties expend much energy arguing whether ad revenue is important to cable, and whether, if now not very important, it is likely to become much more so shortly. Because the quantitative importance of ad revenue is immaterial in the view I take of the matter, I assume throughout, in favor of defendants, that ad revenue is of great and rapidly growing importance to cable.
To whatever extent broadcasters are competitively hobbled by § 534(b)(3)’s implicit ban on ad avails, however, they cannot rely on it to justify the constitutionality of the rest of the Act. The government surely does not have a “substantial interest” in remedying a competitive distortion that arises entirely out of a detail in its own purportedly remedial legislation.
Further, advertising space is simply one form of the currency that changes hands as cable operators negotiate carriage deals with content providers. Just as a home-shopping channel might offer an outright payment or a commission on sales-to every conduit that will carry it (either broadcast station or cable operator), presumably the disadvantaged broadcaster can simply find another way to sweeten the pot when competing for a carriage slot against a cable programmer offering ad avails. At least it could do so if Congress had not simultaneously barred such side payments. See 47 U.S.C. § 535(i). For example, but for the congressional ban on such payments, a broadcaster could offer a cable operator some share of the ad revenue increment expected as a result of carriage on the cable system, and thus compete for space on cable on a par with independent operators. The defendants’ chief expert explicitly acknowledged this parity. See Noll Deposition at 54-55, 104-05. Thus, insofar as broadcast stations suffer a competitive disadvantage arising out of their failure to offer ad avails, it arises entirely out of a provision in the remedy Congress has enacted and cannot possibly constitute the “substantial interest” that Congress was seeking to correct.
Accordingly, defendants’ primary expert concluded — and the record contains no material evidence to the contrary — that with reference to this supposed market imperfection the only genuine difference between an independent program network and a broadcaster was that the weakening of a broadcaster harms non-cable households through reduced revenue and reduced investment in programming. See id. at 105-06. But, as the discussion of the viability of the broadcast industry has shown, the Court believed that the interest in programming for noneable households could serve as a potential justification for must-carry (apart from actual distortions of the competitive market) only if the absence of must-carry would leave broadcast’s capacity to serve the nation’s uncabled 40% seri*778ously impaired, a standard not met by defendants’ evidence.
I would grant summary judgment for plaintiffs on defendants’ theory that cable poses an illegitimate competitive threat to broadcast by virtue of independent cable programmers’ supposed (but plainly mythical) advantage over broadcasters in seeking access to viewers via cable. The facts show that Congress could not have so found.
(b) Broadcasters as conduit competitors with cable
The defendants claim that a cable operator will have an incentive to weaken broadcast competition in general through carriage denials, regardless of whether the broadcaster is replaced with a cable network. See Dertouzos Reply Decl. ¶23. The theory here is that even broadcasters that would otherwise be worthy of carriage in the short term (since their presence would enhance the cable system’s appeal to its subscribers) will be denied carriage in an attempt by cable operators to strangle them. Cut off from the incremental ad revenue, marginal broadcast stations will decline (disabled from investing more in programming) and, at the limit, die. Especially if there is a separate market for local advertising on limited-viewership television (presumably with lower prices to reflect to smaller audience), then destruction of the marginal broadcasters will redound to the benefit of the cable operator. Even in the weaker version, which contemplates carriage denials causing only lower station revenue and therefore fewer program quality enhancements, cable is seen to profit “unfairly” in that its carriage denials induce marginal non-cable viewers to switch to cable, impairing the broadcasters’ competition for ads.
The difficulties with this as a serious justification for must-carry are legion. First, the cable operator must share many of the benefits of this hypothetical policy with competing broadcast stations: so long as they are around, the disappointed viewers of the station denied carriage can still turn to other broadcast stations (not to mention book stores, libraries and unnumbered other sources of diversion and instruction), rather than to cable. Cf. Noll Deposition at 78-79 (“It seems to me that if there’s three UHF independents in a city and one is dropped and two aren’t, the two that aren’t dropped benefit; the one that is dropped is harmed, and I don’t understand how to add that up to be health of the [broadcast] industry.”)
Second, given the defendants’ concession that must-carry is generally of interest solely to the more marginal stations, Defendants’ Fact Response ¶¶ 65, 73, because a cable operator would not dare to incur subscriber wrath by dropping stations as popular as local network affiliates, id. ¶ 62, the likelihood of the policy’s yielding many conversions to cable seems remote.
Further, so far as inflicting fatalities on broadcast stations is concerned, the hypothesized policy depicts the cable operators as absolute Canutes: as the tide of broadcast stations rises, including even stations of the least prosperous class, see Graphs # 3 & # 4, it must take an invincibly obtuse operator to persist in trying to push it back — at least if persistence is costly, as it would be wherever adding the broadcast station would add to net revenues. And there would be such an addition to revenues, in the absence of artificial congressional impediments, whenever the additional revenue expected from carrying the broadcast station — in subscriber fees, broadcast station payments or the yield from ad avails — exceeded those on offer from independent programmers proposing to fill equivalent channel capacity.
Finally, the problem identified appears ubiquitous, so that if it is enough to sustain forced carriage it must be enough to sustain a wide range of forced speech. The Washington Post would presumably gain at the margin from the demise (or starvation) of its competitors; yet, despite the enormous economies of scale that it enjoys (and a huge local market share, cf. Business News Two, Investor’s Business Daily, March 28, 1995, Al), I doubt that Congress could legitimately force the Post to carry whole pages of the Washington Times, with the Times then able to obtain higher ad revenue from advertisers who would see their ads freely circulated to Post readers as well.
*779Nonetheless, because the defendants offer some anecdotal evidence of operator predation (resolution of which might require trial), see, e.g., Plaintiffs’ Fact Response ¶¶ 46-47,1 will assume that summary judgment cannot be granted in plaintiffs’ favor on defendants’ claim that a kind of competitive spite may induce cable operators to deny carriage to some broadcast stations.
IV. Narrow Tailoring
I have found some of the assertions of harm offered by the government to justify must-carry to be either pure conjecture (and largely refuted by the defendants’ own experts and data), and some so self-evident that if the Supreme Court regarded them as adequate it would not have remanded for additional factfinding. In the former category rests the possibility that the future of free television as a whole or even in a particular geographic market is threatened. All parties and the data now appear to agree that no such threat is shown. See sections III.A.1 and III.A4 above. In the latter category are defendants’ contentions that must-carry was intended as a way to (1) maximize broadcast’s wealth in the interest of increased expansion and quality of programming or (2) assist whatever stations might elect it. See sections III.A.2 and III.A.3 above.
It follows from these latter definitions of the harm that must-carry is by definition narrowly tailored, just as the harm itself exists by definition. Thus, confronted with the suggestion that narrower must-carry rules, such as those struck down in Century Communications Corp. v. FCC, 835 F.2d 292 (D.C.Cir.1987), might be a better tailored alternative, the defendants answer, inexorably, that if the Century rules didn’t help as many stations as the 1992 Act (and they didn’t), it follows by definition that they are not viable alternatives. See Government Reply at 40. Similarly, since the defendants posit an interest in enhanced broadcast programming, which as explained above could logically be expected to flow from the revenue enhancements that must-carry would generate, it follows as a matter of logic that any solution that simply gives all would-be suppliers of cable programming an entitlement to carriage for a fee remunerating the cable operator’s costs is inadequate: such a system would contribute less money to broadcast stations’ coffers. Government Reply at 43-45 (dismissing leased access because it would “defeat must-carry’s purpose to the extent that [a broadcaster’s] investment in programming must be curtailed to pay for carriage”). If the defendants’ self-validating concepts of “harm” are indeed what the Court intended us to examine, must-carry fits them so perfectly that, as a matter of definition, it must meet the requirements of narrow tailoring as well.
On the other hand, if I am right that the alleged threat to “access to free television programming for the 40 percent of Americans without cable,” Turner, — U.S. at -, 114 S.Ct. at 2461, was conjectural rather than real, the question whether must-carry constitutes a narrowly tailored remedy for injury to that “access” is moot.
Of the two remaining asserted harms by cable operators to the broadcast industry— those distinct from a claim that broadcast is in general ill health — I have assumed that Congress might reasonably have found a problem to be fixed. First, I assumed that the affiliation of a cable operator with a cable programmer, especially in an environment of rate regulation, could result in an unjustifiable competitive edge for the vertically-integrated firm against unaffiliated programmers, including broadcasters. See section III.B.l above. And I assumed enough evidence that cable operators might conceivably be willing and able to inflict predatory harm on broadcasters that I would not necessarily grant summary judgment in favor of plaintiffs on the issue. See section III.B.2(b) above.
I turn, then, to the narrow tailoring step of the analysis under O’Brien, canvassing the evidence introduced on remand to answer the questions contingently put to us by the Court, examining “the extent to which cable operators will, in fact, be forced to make changes in their current or anticipated programming selections; the degree to which cable programmers will be dropped from cable systems to make room for local broadcasters; and the extent to which cable operators can satisfy their must-carry obligations *780by devoting previously unused channel capacity to the carriage of local broadcasters.” Id. at -, 114 S.Ct. at 2472. I then consider the availability and efficacy of “constitutionally acceptable less restrictive means” of achieving those asserted interests among the legitimate justifications for must-carry that have been proven or assumed real: preventing anticompetitive behavior by vertically-integrated cable firms and preventing predatory behavior against broadcasters by cable in general. Id.
A. Must-carry’s Burden to Operators’ and Non-Broadcast Programmers’ Speech
The record shows that there are currently 128 national and 40 regional cable networks (i.e., programming channels available to cable systems), and the parties agree that the number is expected to grow. Defendants’ Fact Response ¶ 107. It is also undisputed that as of October 1994 an estimated 86% of cable systems had a maximum capacity of less than 54 channels. Id. ¶ 112. An overwhelming majority of cable systems therefore had over a hundred programming sources available (not counting broadcast stations), far exceeding what they could carry. Indeed, a study by plaintiffs found that in October 1994 cable systems serving about two-thirds of the industry’s subscribers had no available, usable channel capacity. Id. ¶ 115. The defendants offer a survey which shows that 51% of cable systems have no excess capacity. Id.; Plaintiffs’ Fact Response ¶ 193. (These surveys are not necessarily in conflict; the first deals with number of subscribers served and the other with the raw number of cable operators affected.)
Must-carry is thus most commonly elected — by defendants’ data, 51% of the time, all else being" equal — in an environment of scarce cable channels. Defendants’ additional data show that in 1993 must-carry stations added as a result of the Cable Act accounted for between 1.4% and 6% of the channels offered by sampled systems, and total must-carry stations accounted for between 7% and 13% of channels. Declaration of John L. Peterman Ex. 1. The average percent, weighted for subscribership, was 2.4% for added stations and 12.28% for total stations. Id. Ex. 2. This forced carriage — whether in the form of a required addition to a cable system (requiring the operator to eliminate programming) or a mandatory retention (preventing the operator from adding new programming) — is a tangible burden on operators’ speech.23
Defendants implicitly claim that the First Amendment harm to a programmer denied carriage in the first instance because of a cable system’s full capacity (partially thanks to must-carry) is somehow more remote than the harm experienced by a carried programmer that is dropped in order to make room for forced carriage of a broadcast station. While the former situation may be more difficult to prove, it is no less a real manifestation of the effects of must-carry. Accordingly, it follows that, except to the extent that there is excess channel capacity (a matter we reexamine below), every benefit to a broadcast station (through being added or being sheltered against removal) is matched by an injury to a cable operator (in lost choice and possibly in lost revenue) and programmer (through being dropped or, thanks to a broadcaster’s must-carry entitlement, not added).
I note various conflicting claims by the parties as to the comparative value of their respective program material, accompanied by disclaimers of the relevance of such data. See, e.g., Plaintiffs’ Facts ¶¶ 9, 176, 177; Defendants’ Fact Response ¶¶ 8, 9. I think the disclaimers are sound. The value measures are necessarily rather crude: as each side *781points out, the fact that a program has a lower rated viewership than another cannot be a basis for congressional preference.24 Accordingly, in the systems of constrained capacity, which by any measure constitute at least half the market, benefits are balanced by burdens — except that congressional choice has supplanted that of cable operators seeking to maximize their net return.
Defendants point to an ancillary benefit of must-carry: some fraction of the revenue enhancement enjoyed by a carried broadcast station will redound to the uncabled 40%’s benefit in the form of additional program investment. Of course a parallel ancillary benefit results from carriage of a cable programmer: just as added revenue both enables a broadcast station to invest more in programming and gives it added incentive to do so, so too must any additional revenue enjoyed by a cable programmer as a result of carriage. Nothing in the record supplies a clear reason for supposing one ancillary benefit more valuable than the other. Except to the extent that broadcast survival is at stake, as it is not, the purpose in such a wealth transfer is elusive.
Even looking simply at the class of programmers forced to “go dark” to make room for new must-carry stations in areas in which they had previously been carried through cable, the harm from must-carry is substantial. For example, Time Warner added 240 broadcast stations to its systems because of must-carry and had to drop programming on 75 channels as a result. Defendants’ Fact Response ¶ 117. This is an obvious and direct burden on the First Amendment rights of both the programmers affected and the cable operator, who had freely contracted with the programmers to air their speech.
As indicated, 49% of cable systems have open capacity. It does not follow, as defendants would have us believe, that such systems suffer no injury from must-carry. Capacity is added at a cost, and one may presume that cable operators add capacity only where they suppose that they can profitably use the capacity. See Noll Dep. 258-59 (excess capacity is a short-run phenomenon, arising from fact that new capacity is added in “discrete lumps” that are not instantly filled). (When a retailer holds goods in storage, or a manufacturer holds parts in inventory, their existence presumably contributes to the business despite the absence of immediate active use.) Even for a system that generally keeps a few channels open, presumably for flexibility in responding to carriage opportunities, must-carry is a constraint, inevitably hastening the day when it must invest in additional capacity. Thus, defendants’ evidence that the must-carry duty now embraces only a weighted average of 12.28% of the capacity of cable systems, and that that percentage will likely shrink as capacity is added, see, e.g., Defendants’ Fact Response ¶¶ 196-201, in no way answers the point that any station carried involuntarily blocks the cable operator’s intended use of the channel, forcing it to make changes in its anticipated, if not current, programming selections. If in the future fiber optics or compression truly make functionally limitless channel capabilities available at no extra marginal cost, must-carry’s burden would certainly be less along this dimension. But that day has simply not arrived.
The channel positioning requirements of the Cable Act, 47 U.S.C. §§ 534(b)(6) and 535(g)(5), further demonstrate the intrusion into operators’ carriage decisions and programmers’ speech. For example, the Act offers commercial stations electing must-carry a channel position either the same as their on-air positions or on positions previously occupied on July 19,1985, or January 1,1992, whichever they prefer. Since cable systems’ capacities were much lower ten years ago, this provision tends to permit stations whose forced carriage in 1985 had necessarily given them a coveted low-numbered channel slot to insist upon that slot once again. The defendants put great weight on channel position within a system. See, e.g., Government *782Brief at 18-22. Channel position, they claim, is of great value because viewers apparently tend to “graze” in the lower channels, not even bothering to look among higher-numbered ones (a zone known as “Siberia,” Defendants’ Record Vol. I.R. Ex. 85 CR 11503-04). WTOG, which is on the air as channel 44, thus decries as an adverse repositioning a cable system’s decision to demote it from channel 9 to channel 25 — even though that is clearly preferable to its over-the-air position. Defendants’ Record Vol. I.R. Ex. 83 CR 11478.
Moreover, according to one UHF station president, “[e]xisting cable technology prevents many [cable] subscribers from accessing cable channels above 13 without a converter.” Defendants’ Record Vol. I.R. Ex. 85 CR 11503-04. For those cable subscribers, forced repositioning of a cable programmer’s channel to a higher-numbered station will simply eliminate the channel from view. Those same viewers can use their television sets to receive broadcast channels above 13. Affidavit of David J. Large (“Large Affidavit”) ¶ 15; Defendants’ Record Vol. I.R. Ex. 85 CR 11504. In such cases, then, a UHF station can reach the viewer’s set twice— through broadcast on its own over-the-air channel and through forced carriage on a lower cable channel — while the demoted cable channel will be unattainable.
The congressional mandate on repositioning casts the defendants’ contentions about the projected future expansion of cable channel capacity in a special — and dim — light. Even the most optimistic claims of expansion must, by defendants’ own accounting, be tempered by the fact that as cable expands, the only place it can grow is in the wastes of post-channel-13 Siberia. Large tracts of the lower “grazing areas,” and any real or imagined boost to ratings that positioning there provides, are claimed by the government for allotment to the broadcasters — even ones that cannot obtain such favorable positioning within the government’s own broadcast spectrum allocation — and to that degree are barred from use for cable programming.
In sum, on systems of clearly constrained capacity the benefits that must-carry affords broadcasters are matched by burdens on cable operators and programmers. In systems whose capacity is not fully occupied by active channels, must-carry interferes with the operators’ purpose in maintaining channels available for use. Although increasing channel capacity will over the years tend to diminish the value of the losses inflicted on cable operators and programmers, it will never do so entirely, because the marginal opportunity cost of channels will presumably never fall to zero and because broadcast stations’ repositioning entitlements will force alternative cable programming into high-numbered, less valued channels.
B. Less Restrictive Alternatives
Assessing the availability of less restrictive alternatives to must-carry helps determine the reasonableness of the “fit” between forced carriage and the fulfillment of any substantial interests established by defendants. Because in the first step of the analysis under O’Brien I have only assumed two such interests — the interests in preventing anticompetitive behavior by vertically-integrated cable firms and in preventing predatory behavior against broadcasters by cable in general — I focus mainly on fitness to those purposes. Occasionally, however, I will make reference to those interests whose substantiality the record does not support or that the Court’s Turner decision implicitly rejected. I consider three possibilities in this light — genuinely non-discriminatory entitlements to carriage at regulated rates, the “A/B” or input selector switch, and the rules struck down by the D.C. Circuit in Century Communications Corp. v. FCC, 835 F.2d 292 (D.C.Cir.1987), clarified, 837 F.2d 517 (D.C.Cir.1988).
1. Leased, non-discriminatory access
One cannot correct an unlevel playing field by retilting the field to the benefit of some previously-disadvantaged players — broadcasters — at the expense of others who were already more severely disadvantaged — independent programmers. Because must-carry suffers from this defect, it is in virtually every respect mismatched to the concern that vertically integrated cable bottlenecks *783will distort program selection choices to the detriment of viewers.
First, the remedy is substantially overinclusive, in that it applies to all cablecasters, including ones completely free of vertical integration. Cf. Simon & Schuster, Inc. v. Members of the New York State Crime Victims Board, et al., 502 U.S. 105, 121-22 & note, 112 S.Ct. 501, 511 & note, 116 L.Ed.2d 476 (statute was overinclusive even under content-neutral scrutiny because a substantial portion of its burden on speech did not serve to advance government goals). As even defendants claim only that as of 1994 operators serving more than 70% of cable subscribers held vertical interests in cable programmers, Government Reply at 30-31, Plaintiffs’ Fact Response ¶¶ 88-89, clearly those serving 30% do not. For the class of non-vertically-integrated cable operators, must-carry levels a playing field that was not tilted to begin with. Moreover, where the affiliated program supplier provides only a limited line, e.g., programming for one channel, the putative foreclosure effect appears miniscule. A “solution” that preempts 20 channels in order to provide “fair competition” on one plainly burdens more speech than is necessary.25
Second, although the structural defect within vertically-integrated cable firms disadvantages independent programmers in precisely the way it disadvantages broadcasters, see Noll Decl. ¶ 19; Noll Dep. at 54-55,104-06; section III.B.2(a) above, must-carry singles out only broadcast for forced carriage. This defect of the remedy is not merely a matter of Congress exercising its authority to solve a problem “one step at a time,” cf. Williamson v. Lee Optical, 348 U.S. 483, 489, 75 S.Ct. 461, 465, 99 L.Ed. 563 (1955), and to refrain from burdening additional classes of speech. Rather, in two ways it undercuts any claim that must-carry is intended as a remedy for the stated problem. First, the limitation of the must-carry right to broadcasters renders the right "ineffective or remote” as a solution to the problem. Central Hudson Gas & Elec. Corp. v. Pub. Serv. Comm’n, 447 U.S. 557, 564, 100 S.Ct. 2343, 2350, 65 L.Ed.2d 341 (1980); Edenfield v. Fane, 507 U.S. 761, -, 113 S.Ct. 1792, 1799, 123 L.Ed.2d 543 (1993); see also FCC v. League of Women Voters, 468 U.S. 364, 396, 104 S.Ct. 3106, 3126, 82 L.Ed.2d 278 (1984); Florida Star v. B.J.F., 491 U.S. 524, 540, 109 S.Ct. 2603, 2613, 105 L.Ed.2d 443 (1989) (government “must demonstrate its commitment to advancing [its] interest by applying its prohibition evenhandedly”). Second, the privileged status of broadcasters here comes substantially at the expense of the other injured parties, the unaffiliated programmers, narrowing their opportunities to compete for channels. See Turner, — U.S. at -, 114 S.Ct. at 2456. Yet the most obvious distinction between broadcasters and unaffiliated programmers is one that argues for less protection for broadcasters: they, in contrast to independent programmers, have their own access to viewers — at a minimum to the 40% who lack cable (even if one is assuming a station denied carriage by all cable systems in its viewing area).26
Finally, the must-carry remedy is both more burdensome to speech and less fit to advance Congress’s expressed competition-protection purposes than are standard regulatory tools for assuring accommodation of would-be users of a bottleneck. In industry after industry — for example, interstate transmission of natural gas, interstate transmission of electricity, local area telephone networks — federal and state legislatures and regulatory agencies have sought to assure access for would-be users on equal terms, and at prices at least allowing the bottleneck owner to recover the associated costs. See, e.g., United States v. Terminal Railroad Ass’n, 224 U.S. 383, 411-12, 32 S.Ct. 507, 516, 56 L.Ed. 810 (1912) (permitting access on equal terms for all railroad companies to railroad bottleneck controlled by a subset); Federal Energy Regulatory Commission Or*784der No. 636, Pipeline Service Obligations and Revisions to Regulations Governing Self-Implementing Transportation, etc., 57 Fed.Reg. 13267 (April 16, 1992) (establishing terms for interstate pipelines’ transportation of others’ gas); FERC Notice of Proposed Rulemaking, Promoting Wholesale Competition Through Open-Access Non-Discriminatory Transportation Service by Public Utilities, etc., 60 Fed.Reg. 17662 (April 7, 1995); Federal Power Act §§ 211 & 212, as amended, 16 U.S.C. §§ 824j, 824k (Supp.1955) (providing limited authorization for orders to “wheel” electricity); MCI Telecommunications Corp. v. FCC, 561 F.2d 365 (D.C.Cir.1977) (“Execunet /”); MCI Telecommunications Corp. v. FCC, 580 F.2d 590 (D.C.Cir.1978) (“Execunet II”); United States v. AT & T, 552 F.Supp. 131, 227 (D.D.C.1982) (provision of Modified Final Judgment assuring non-discriminatory access of interexchange carriers to local loop).
Such prices are critical if the intervention is actually to correct the distortion without creating others that do equal or greater damage. If there is no charge, or if the charge is set too low, there is a risk that the bottleneck operator will pass a higher proportion of the protected outsiders’ material through than would an operator under perfectly competitive conditions. See Noll Dep. 183-85 (acknowledging that must-carry could result in carriage of more broadcast stations than in perfectly competitive environment); see also William J. Baumol & J. Gregory Sidak, Toward Competition in Local Telephony 73 & ch. 7 (“The Pricing of Inputs Sold to Competitors”) (1994). Conversely, too high a price would disfavor the outsiders. While there are sharp disputes as to exactly how such charges should be calculated, see, e.g., William J. Baumol & J. Gregory Sidak, The Pricing of Inputs Sold to Competitors, 11 Yale J.Reg. 171 (1994); William B. Tye, Response, id. at 203; Aired E. Kahn & William E. Taylor, Comment, id. at 225, I am unaware of any economist arguing that singling out a special set of non-affiliated upstream suppliers and granting them access at a price of zero is a “solution” to the bottleneck problem. The universal solution in such mundane areas as gas, electricity, and wire telecommunications is a non-discriminatory entitlement to access at a price set by an administrative agency. Indeed, in section 9 of the 1992 Cable Act (amending section 612 of the 1984 Act) Congress adopted a comparable provision for cable, setting aside a portion of each operator’s channels for mandatory “leasing.” 47 U.S.C. § 532. Regardless of whether defendants’ complaints about the adequacy of that specific remedy are correct, see Government Opposition at 92-93, the type of remedy used for gas, electricity and wire telecommunications appears to solve any bottleneck problems of broadcasters in a way that creates no ancillary burdens on others competing for bottleneck access. And no defendant even suggests the contrary.
In fact, the government’s response to leased access is precisely that such access presupposes payment by broadcasters to cable — “one of the very evils that Congress noted.” Id. But obviously carriage has some value. While the government may decry an attempt by cable operators to overcharge a programmer for carriage, the claims of anticompetitive or “unfair” behavior provide no reason to treat any charge— including one worked out through the processes described in comparable situations above — as an “evil.”
If carriage generates for broadcasters the enormous incremental revenues that defendants claim, it is especially odd to find them disparaging the idea that they should pay any of the costs of access to this bonanza. In describing the value of carriage to broadcast stations, for instance, defendants point to evidence of a cable operator’s estimate that for a particular broadcaster “for every 5% drop in cable viewers, this station would lose $1,480,000 in gross revenues and $1,258,000 in operating income.” Defendants’ Fact Response ¶ 54. I suspect the station in question is not typical of those allegedly jeopardized by the absence of must-carry, but the point is the same for all stations. Neither in terms of equity nor of economic efficiency have defendants offered a reason why broadcast stations should not bear any of the costs — or pay any of the value — of the access that they identify as valuable.
*785Thus the government fundamentally errs when it claims that mandatory free carriage for broadcasters serves consumers’ interests as measured by the standards of a free market. The Government says:
It is cable operators who deny access by significant numbers of local broadcast stations to a significant share of their market, thereby disabling that market from ever reaching a ‘verdict.’ The must-carry rules address this reality, and are premised on the principle that, after free access to the marketplace is restored, fair competition can begin.
Government Opposition at 41. But granting broadcasters free (as in “no charge”) access to cable’s subscribers — and thereby denying access to independent programmers — certainly does not enhance free (as in “free and open”) competition between broadcasters’ programming and that of independent programmers. Ultimately, in the absence of an overall threat to broadcast’s survival, the salient difference between independent programmers and broadcasters is that the latter enjoy use of a valuable asset given them by the government — a broadcast license — that the former lack. I cannot see how principles of “fair competition” are enhanced by a rule under which the broadcast license, itself a government subsidy, becomes an admissions pass to an additional government-enforced subsidy, paid for by cable operators and independent programmers.
Accordingly, I would conclude that must-carry violates the First Amendment requirement to use the least restrictive alternative, as that has been understood since Ward v. Rock Against Racism, 491 U.S. 781, 109 S.Ct. 2746, 105 L.Ed.2d 661 (1989), by adopting a mandate of completely uncompensated carriage in favor of a privileged class of users, at the expense of other speakers whose access to viewers is subject to more severe limits.
The analysis is parallel for public television. Access to cable subscribers who have abandoned their antennae is undoubtedly valuable for public stations, and I have already found support for a congressional fear of a risk of discrimination against broadcasters and other unaffiliated program suppliers. But just as commercial broadcasters could use part of the anticipated extra revenue from nondiscriminatory access at a regulated rate, so too, public stations could pay for the access with a part of the extra contributions it is expected to yield. The only reason offered by public television for more favorable treatment — apart from content-based paeans to the quality of its programming, see Public Broadcasters Defendant-Intervenors’ Memorandum of Points and Authorities in Support of Motion for Summary Judgment at 15-18 — is that the public is already subsidizing public television and is entitled to protect its investment. See id. at 19-21. But public television advocates point us to no precedent for the proposition that a watered down version of least-restrictive-alternative analysis is to be applied to a case of coerced speech simply because the stated purpose is to enhance the value of a previously granted public subsidy.
I close by acknowledging what is obvious about the governmental purposes not based on “unfair” competition. If the defendants are right that the government has a “substantial interest” in maximizing the revenue or net profits of broadcasters, or maximizing the growth rate of UHF independents, then any charge, no matter how legitimate and no matter how parallel to permitted charges in other bottleneck situations, will fail to advance the interest as well as a free, fiat entitlement. I have already stated my belief that the Court did not endorse those goals as substantial government interests. But for the stated goals of preventing “unfair competition,” in any of its alleged manifestations, non-diseriminatory access at agency-regulated fees burdens far less speech and fits— rather than fights — the stated goals.27
2. A/B or “input selector” switches
An “A/B” or input selector switch, despite the slightly forbidding name, is simply a *786switch with one position for cable and one position for broadcast. Coupled with whatever antenna is necessary (e.g., the one in place before the viewer subscribed to cable), such a switch permits the viewer to switch between cable’s offerings and all the television programming that would have been available to the viewer if cable had never been installed.
All parties have pitched their arguments about the use of such switches in terms of whether they qualify as a “less restrictive alternative” to must-carry. I frankly doubt this classification. The availability of A/B or input selector switches is simply part of the landscape. If they enable cable subscribers to reach programs they wish to see, as plaintiffs argue, that is simply a fact that tends to undermine the claim that the continuation of over-the-air television is in any way in jeopardy or that cable operators have a “bottleneck” over provision of video services to cable subscribers (a foundation of any claim of anticompetitive behavior).28 If defendants are right in their assertion that the use of such switches is infeasible, that fact merely supports their claims as to the fate of marginal broadcast stations and the existence of bottleneck distortions in the market. Nonetheless, as both parties treat these switches as some sort of “alternative,” I address them here in that light.
A cable subscriber is able to retain access to broadcast television if willing (1) to pay the $10 or less required for the A/B or input selector switch (in some cases the subscriber receives it as part of a TV’s remote control), Large Affidavit ¶22, (2) to use the switch, and (3) to retain whatever antenna is necessary to pick up over-the-air signals.
The first barrier is apparently lessening. The record shows that most cable-ready TV receivers now on the market have built-in input selector switches operable by remote control. Defendants’ Fact Response ¶ 154. The record does not disclose, however, just what portion of the existing stock of television receivers has this advantage. Id. In any event, it would seem that anyone who could afford cable could also afford the $10 for an A/B switch for use with an old television set.
The defendants claim that A/B switches do little to ease cable’s monopoly in the houses of its subscribers because consumers find them “unacceptable.” “Most telling,” the government says in its evaluation of A/B switches, is that “79% of cable viewers stated that the A/B switch was inconvenient for the plain and simple reason that they ‘had to get up’ to use it.” Government Reply at 41.29
“Telling” indeed: the Government’s point tells us that consumers are willing and able to use an A/B switch if they believe it will provide access to a program that is worth “get[ing] up” from the couch for, once to switch to broadcast, once to return to cable. (We put aside those with input selectors in their remotes, who need not even get up.) Compare Century, 835 F.2d at 302 (rejecting the FCC’s “sluggish profile of the American consumer” as inconsistent with fact that “even costly items like the video-cassette recorder, the cordless telephone, the compact disc-player and the home computer have spread like wildfire”).
The 1992 Cable Act in fact precipitated a controlled experiment in consumer resource*787fulness. In late 1993 the three major television network local affiliates were unable to reach agreement on retransmission consent with a TCI cable system in Corpus Christi, Texas, resulting in a three-month span in which ABC, CBS, and NBC were all absent from cable. In an attempt to placate irate cable subscribers, TCI provided A/B switches so that the subscribers could switch from cable to broadcast and back again. See Defendants’ Additional Evidence Vol. VII.M Ex. 291; Large Affidavit ¶ 71.
Before the dispute the networks collectively captured a 70% share of the total cable audience; after a month off cable (and with the distribution of A/B switches), they had a share of 67-69% of the total cable audience. Defendants’ Fact Response ¶ 161; Large Aff. ¶72; Defendants’ Additional Evidence Vol. VII.M Ex. 291.
Defendants note that nearly 70% of sampled cable subscribers in Corpus Christi said the switches were “inconvenient” to use. Defendants’ Fact Response ¶ 161. It is from this sample that 79% said the switches were inconvenient because they “had to get up” to change them. Defendants’ Additional Evidence Vol. VII.M. Ex. 291. But government-mandated carriage seems a rather astonishing solution to a problem now redefined as consumers’ occasional need to “get up.”
The Corpus Christi episode lends credence to plaintiffs’ contentions that the primary reason viewers choose not to use A/B switches is that cable already carries the broadcast stations that viewers regard as worth switching to. “Defendants concede as much when they address the fact that subscribers to direct broadcast satellite, or DBS, must use an A/B switch to retain access to any local broadcast stations because the DBS feed does not include them. (DBS has joined over-the-air broadcast and cable as another way of receiving video signals.) The government claims that ‘[u]se of A/B switches’ by DBS and MMDS [yet another video delivery system] subscribers involves the completely different situation where the consumer must resort to A/B switches to receive virtually any broadcast stations whatsoever.” Defendants’ Fact Response ¶ 158. This appears to be no more than an ornate way of acknowledging that consumers will use A/B switches whenever they want to — i.e., whenever they are aware of a non-carried station that is worth a few dollars’ capital investment.and the occasional flick of a switch. Defendants do not contend that DBS subscribers — or Corpus Christi residents — are more handy or spry than Americans generally.
Defendants point out that use of an A/B switch is successful only to the extent that a suitable antenna will pick up the desired signal, and observe that antennae may costs as much as $300. (There is no suggestion that that is a common price, or anywhere near average.) Yet the uncabled J0% of America, which in this litigation is often presumed to be unable to afford cable, somehow manages the burdens of antenna ownership, as do DBS subscribers, who alone amounted to 1.53 million people in 1994. Large Affidavit ¶ 73; Defendants’ Fact Response ¶ 158. Interestingly, it appears that the desire to secure local broadcast signals that are not supplied by cable, DBS or some alternative service is powerful and rising: from 1993 to 1994 there was a 51% increase in the sales of indoor antennas (from 3.5 million units to 5.3 million units). Affidavit of Joseph Stern ¶ 33.
Congress made specific findings regarding A/B or input selector switches in the 1992 Cable Act. Section 2(a)(17) states that
[ (a) ] [consumers who subscribe to cable television often do so to obtain local broadcast signals which they otherwise would not be able to receive, or to obtain improved signals. [ (b) ] Most subscribers to cable television systems do not or cannot maintain antennas to receive broadcast television services, [ (c) ] do not have input selector switches to convert from a cable to antenna reception system, or [ (d) ] cannot otherwise receive broadcast television services.
Section 2(a)(17) (emphasis added). Section 2(a)(18) states that
[ (e) ] [c]able television systems often are the single most efficient distribution system for television programming. [ (f) ] A government mandate for a substantial societal investment in alternative distribution *788systems for cable subscribers, such as the ‘A/B’ input selector antenna system, is not an enduring or feasible method of distribution and is not in the public interest.
Section 2(a)(18). Several of the drawbacks mentioned — items (b) and (c) — are simply conclusory assertions of the supposedly insuperable barriers posed by the need for switches and antennae. But these are belied by the Corpus Christi episode and by the behavior of DBS subscribers. Neither switch nor antenna is so mysterious, costly or unworkable as to be beyond the means of a private citizen who values access to a program at the price of a diminutive capital investment and the ongoing cost of pushing the switch when necessary.
The remainder of the findings — see items (a), (d), and (e)30 — appear to reject A/B switches in large part on the ground that cable in many cases enhances broadcast stations’ signals or even brings them to areas they could not otherwise reach. Cf. Noll Declaration ¶ 40; Dertouzous Declaration ¶31; Government Opposition at 91. But neither the enhancement of television options for cable subscribers, nor extending broadcast’s reach to cable subscribers in otherwise signal-deprived areas, was embraced by the Supreme Court in Turner as an interest substantial enough to justify must-carry. See section III.A.2 above.
In fact, use of cable to enhance (or create) a signal in previously-signal-deprived areas may affirmatively injure the uncabled 40%. Again, obviously only cable subscribers would directly enjoy the enhanced (or new) signal. Further, as forced carriage would reduce a station’s potential audience to be reached over-the-air, it would pari passu reduce the station’s incentive to invest in enhancement of its over-the-air signal (say, by buttressing its transmitter).
Moreover, while a cable operator in areas where broadcast signals are weak or nonexistent may have a “monopoly” not only over cable but also over television services more generally, defendants’ claims about cable operators’ supposed motives to impoverish or destroy broadcast stations for anticompetitive reasons, see section III.B.2(b) above, cannot apply in such areas: cable operators have no motive to destroy a “competitor” that is unable to compete.
The insufficiencies of A/B switches in weak-signal areas, nonetheless, make them, in such areas, an inadequate alternative solution to Congress’s stated concern that cable’s “bottleneck” powers might deny consumers adequate access to diverse programming from unaffiliated programmers. See sections III.B.l and III.B.2(a) above. But the defendants have made no effort to quantify the weak-signal areas, which from all that appears may represent a very small fraction of the country. Accordingly, as the 1992 Cable Act imposes its burdens on speech throughout the country, while broadcast’s own access to consumers via A/B switches adequately counters cable’s alleged anticompetitive tendency to foreclose broadcast programming everywhere except in weak-signal areas, it burdens more speech than is necessary even with respect to those purposes.31
3. The Century Rules
The plaintiffs appeal to the previously-struck-down Century rules, 835 F.2d 292, which required fewer forced carriages, as less restrictive. As already mentioned, the defendants agree — and thus urge that they are less effective. For the valid and real purposes that one can identify, however, any *789carriage scheme which is by fiat (1) guaranteed free to the beneficiaries and (2) helpful to only one class of aggrieved parties at the expense of a more vulnerable class, suffers from a similarly fatal mismatch. Less encompassing rules such as those in Century may reduce the current must-carry’s distortive effects, but they still do not amount to a reasonably tailored fit to the problem when there are alternatives available that do not burden speech at all, such as input selector switches, or that at least do not increase the hazards of a class of speakers already suffering from greater bottleneck foreclosure than the legislatively privileged class.
V. Conclusion
Under the standard of Turner, I would find sections 4 and 5 of the Cable Act, 47 U.S.C. §§ 534-535, to be unconstitutional. Defendants have not argued that the special entitlement for “low power” stations, 47 U.S.C. § 534(c); see also Turner, — U.S. at - n. 6, 114 S.Ct. at 2460 n. 6, is severable from the rest of § 534, nor does such a contention seem plausible. I therefore would not reach the issue of whether the low power provisions are content-neutral or content-based, and whether, viewed in isolation, they would be constitutional. Nor would I reach either Time Warner’s takings claim or the claim of Atlanta Interfaith Broadcasters, Inc., that the must-carry provisions are invalid as applied to “predominantly religious television stations.”
Accordingly, I believe plaintiffs’ motion for summary judgment should be granted and defendants’ motion for summary judgment should be denied.
*790[[Image here]]
. 819 F.Supp. 32 (D.D.C.1993).
. As there are 209 "Areas of Dominant Influence,” i.e., non-overlapping geographic areas that define television markets, and over 11,000 cable systems, the typical station would have *757approximately 53 cable systems operating in its "ADI." See All Parties’ Joint Statement of Undisputed Facts ¶¶ 39-45, 136; Defendants’ Response to Plaintiffs' Joint Statement of Undisputed Facts Established by the Record ¶ 16.
. Declaration of Roger G. Noll ¶ 9.
. Some stations — typically popular network affiliates — invoke rights that the Cable Act offers broadcasters as an alternative to forced carriage, namely, to forbid cable carriage until some form of payment from the cable system in question is negotiated. 47 U.S.C. § 325(b). Divisions between stations electing such “retransmission consent" and those electing must-carry, however, are blurry: because a station's coverage area usually includes several cable providers, see n. 1 above, a station can insist on retransmission consent from some cable systems and must-carry on others. See Time Warner Entertainment Company, L.P.'s Memorandum of Law in Opposition to Defendants' and Intervenors' Summary Judgment Motions and in Further Support of Its Summary Judgment Motion at 13-15; Corrected Memorandum of the United States and the Federal Communications Commission in Opposition to Plaintiffs' Motions for Summary Judgment at 10.
. This was not the defendants' original position before this court. Compare, e.g., November 24, 1992 NAB Corrected Memorandum in Opposition to Plaintiffs' Motion for Preliminary Injunction at 14 (“risks of anticompetitive harm are so substantial, Congress concluded, that 'the economic viability of free local broadcast television and its ability to originate quality local programming will be seriously jeopardized,’ absent remedial legislation") (quoting § 2(a)(16) of the 1992 Cable Act); December 2, 1992, Memorandum of Intervenor-Defendant Applicant National Association of Broadcasters in Opposition to Motions for a Temporary Restraining Order at 10-11 (cable's carriage behavior "placed broadcasting's future in jeopardy” and "could deprive the 40% of American households without cable of the opportunity to receive video programming at all”), with NAB Opposition at 18 ("Must-carry was not enacted to prevent the disappearance of free local broadcast television”); Government Opposition at 7 ("Congress did not contemplate, as plaintiffs contend, that the entire broadcast industry is in jeopardy”).
. "Advertising Revenues per Television Household: A Market by Market Analysis,” a 1991 report by the National Association of Broadcasters, reports a "healthy” 20.6% increase in total net advertising revenues for broadcast between 1986 and 1991. Time Warner Ex. 19. It thus appears that if one took into account whatever deductions are made in calculating net ad revenues — evidently expenses such as commissions that broadcast incurred to obtain ad sales, see Time Warner Ex. 14 n. 3 — the picture of broadcast's ad revenue would look still brighter.
. The D.C. Circuit court of appeals struck down the regulations in July 1985, in Quincy Cable TV, Inc. v. FCC, 768 F.2d 1434 (D.C.Cir.1985).
. See Century Communications Corp. v. FCC, 835 F.2d 292 (D.C.Cir.1987), clarified, 837 F.2d 517 (D.C.Cir.1988).
.I am not saying Congress may not act prophylactically, which of course it may. I am only saying that where the evidence shows a threat to be neither present or detectably on the horizon, Congress may not simply obviate the lack of evidence by claiming that all the relevant parties are putting on an act to conceal the threat.
. The year 1987 (meaning 12/31/86 to 12/31/87) was a year of great growth — 72 stations. It is also a period half without must-carry and half with. If the growth can somehow be attributed to the must-carry during the latter half of 1987 or the must-carry prevailing in 1985 and earlier, then absence of must-carry perhaps is responsible for the reduced growth rate described in the text. The parties do not give the court enough detail on stations’ gestation periods, etc., from which one could infer likely time lags between a change in must-carry status and an effect on additions of new stations.
. Defendants use slightly different numbers in discussing the supposedly stunted growth rate, *766see Government Brief at 36, but refer only to three years, 1986, 1990 and 1991, saying that net additional stations in each year were 44, 13 and 11. As 1986 was a year without must-carry, this shows a decline only on the assumption of a long gestation period for stations, and, again, the court is not given enough detail to resolve that. Even accepting the defendants' numbers and apparent attributions of particular years, the data still show significant net growth.
. Plaintiffs also call attention to a 1989 NAB memo which acknowledges that "the future of free TV is excellent," though as a maturing industry, “instead of stations making 50% and 60% profit margins, we are seeing them drop to 30% to 35% margins. Many other mature businesses would love to have these margins.” Time Warner Ex. 4.
. Of course, with or without cable, broadcast must have a natural saturation point — a point where marginal firms earn only a "normal profit,” i.e., a rate of return sufficient to attract capital from alternative investments, all factors (such as risk) being held constant. There is no evidence in the record — at least none identified as such — from which one might infer where that point might be, as an absolute matter or in relation to the number of channels prescribed by the FCC in the rulemaking on which defendants rely. Nor has any party directed the court's attention to anything in the record from which they suggest one might infer that the economic saturation point is greater or smaller than the number of stations made possible by Commission rule.
. One could use similar reasoning to attempt to justify a government requirement that cable operators, at their own expense, print and distribute inserts to accompany their subscribers’ monthly bills, with the content to be determined by local UHF stations. Presumably such stations could use the inserts to advertise themselves, or simply sell the space to other companies for their own generic advertising, coupons, etc.
. While this might map roughly to the subset of stations claimed to be in special financial trouble, see III.A.l.(c) above, it is not an exact proxy for such stations. According to an NAB study, over 75% of all network affiliates elected must-carry on at least one cable system. Affidavit of Gregory M. Loss, Ex. .21 at 5009947. At the same time, the NAB reported that over 90% of network affiliates said they would exercise their retransmission consent rights as to at least one cable system (presumably a system different from one on which must-carry was elected; see n. 3, above). Time Warner Ex. 22.
. The defendants do not advance this as an interpretation of Turner’s mandate, but I examine it nonetheless. In fact, the defendants appear to repudiate such an interpretation, since they claim that carriage of a broadcast station by a cable system is a significant government interest without regard to the number of broadcast stations already carried by that system, or already flourishing in that market. See Defendants’ Fact Response ¶ 11.2.
. The example assumes constrained capacity, such that it would clearly be unprofitable for the operator to run both sets of programming. Thus, a cost of using the affiliate's material is loss of the opportunity to use the non-affiliate’s profitably, and vice versa.
. In contrast, the programming division of the integrated operator/programmer need not be profitable on a stand-alone basis, as long as its net contribution to the integrated firm’s revenue (net of all costs, including risk and the cost of capital) is positive.
.The tradition is breaking down, in that broadcasters have found other sources of revenue. As discussed earlier, under the retransmission con*774sent provisions of the 1992 Cable Act, 47 U.S.C. § 325(b), broadcasters are able to sell the rights to their signals to cable operators, and it is undisputed here that over 80% of all commercial broadcast stations (as opposed to 90% of network affiliates, see n. 14 above,) have elected to do so on at least one cable system. Defendants' Fact Response ¶¶3-5.
. As before, I assume that the operator cannot use both programs, so that use of the outsider’s program costs the operator 20 in foregone ad revenue.
The Cable Act added a complication by requiring that a must-cany station's material be used without change and by barring any payment to the cable operator for carriage. See 47 U.S.C. §§ 543(b)(3), 535(i); see also part III.B.2(a) below. But as Congress cannot invoke a problem created by its "remedy" as a justification for the remedy, these complications are not pertinent.
. Spulber goes on to criticize the theory, which fails to explain why the firm should use the profits derived from deception of regulators for the purpose of predation in the unregulated market if predation there was not otherwise promising. See Spulber, supra, 12 Yale J.Reg. at 60-61. A possible answer is that the otherwise improbable behavior in the adjacent market might be necessary in order to be consistent with the practices engaged in to hoodwink the regulators. In any event, though the theory may be "discredited," as Spulber says, id., I assume that Congress could constitutionally act- on what was until recently the received wisdom.
. Use of must-carry as a means of offsetting the unintended consequences of rate regulation is quite different from using elements of the must-carry remedy itself to justify must-carry; the rate regulation has an independent justification. See Time Warner v. FCC, 56 F.3d 151 (D.C.Cir.1995).
. Of course, to the extent that cable ad revenues are invoked by the FCC as a justification for holding back subscriber rates, the apparent distinction would dissolve.
. I pause to note defendants' argument that the numbers show that must-carry adds as a result of the Cable Act amount to 1.2% of channels and that total must-carry stations amount to 7.2%. See Defendants' Joint Statement of Undisputed Facts ¶¶ 150-53. These figures are arrived at by taking estimates of must-cariy elections and dividing into the nationwide total number of channels. Id. The resulting percentages are meaningless as measures of the foreclosure effects of must-carry; dozens of vacant channels on a handful of several-hundred-channel systems are obviously of no use to a constrained system of 50 channels that is forced to carry, for example, six must-carry stations (slightly under the weighted average). I am surprised that defendants’ counsel offer us such misleading data.
. Of course in a trial on defendants' claims that cable operators often dropped stations for predatory reasons, an operator's judgment about the likely popularity of cable programming selected instead would surely be a defense. Because I have not assumed summaiy judgment for plaintiffs on that issue, however, I do not ultimately confront the context where such a popularity comparison would be pertinent.
. To the extent that defendants' hypothesis of cable operator predation against broadcast stations has merit, as we assumed in section III. B.2(b) above, this finding of overinclusiveness is inapplicable.
. Broadcasters also have access to most of the 60% on cable, for almost all of them can get broadcast signals simply by maintaining their antenna and operating a garden-variety input-selection switch. See section IV.B.2 below.
. Even if one focused solely on the supposed purpose of preventing cable operator predation against broadcasters, see section IV.B.2(b), a nondiscriminatoiy access entitlement at regulated rates fully satisfies the purpose with no ancillary burdening of independent programmers speech opportunities and less burden on those of cable operators.
. The FCC itself pointed out, when promulgating the Century rules, that “once cable subscribers become accustomed to using off-the-air [i.e. over-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." 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) ¶ 163; see also id. ¶ 121 (“[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, ... [and] has caused many subscribers to perceive that there is no need to install or maintain the capability to receive broadcast signals off-the-air.”)
. Actually, the 79% appears to be 79% of the 68% of A/B switch users in Corpus Christi who found the A/B switch “fairly” or "very" inconvenient. See Defendants' Additional Evidence Vol. VII.M No. 291. See immediately below for discussion of the Corpus Christi episode.
. Item (f), to the extent it does not simply repeat earlier assertions, assumes that A/B or input selector switches, and antennae, would become available only by means of some sort of massive government mandate. A survey of the evidence indicates that consumers themselves are quite capable of adopting these measures when they think the extra programs are worthwhile. If Congress feared that consumers were insufficiently aware of such switches, however, it could devise a promotional solution tailored to that end.
. Of course, apart from this ground for granting summary judgment in favor of plaintiffs, section IV.B.l provides an alternative ground: non-discriminatory access for all at regulated rates can meet all the defendants’ bottleneck concerns, see section III.B generally, with less burden on speech and without new, offsetting distortions against programmers — the independents — -who operate at a considerable disadvantage vis-a-vis broadcasters.