OPINION OF THE COURT
AMBRO, Circuit Judge.Table of Contents
I. Background .... 382
A. The 1934 Communications Act and Early Broadcast Ownership Regulation. 382
B Deregulation Initiatives 383
C. The Telecommunications Act of 1996 .. 384
D. Regulatory Review Since 1996.. 384
E. The Commission’s 2003 Report and Order ... 386
F The Order’s Modification of Broadcast Media Ownership Rules ... 386
1. Local Television Ownership ... 386
2. Local Radio Ownership ... 387
3 & 4. Newspaper/Broadcast and Radio/Television Cross-Ownership .. 387
5. National Television Ownership... 388
6. Dual Network Rule ... 388
G. Procedural History of the Current Appeals ... 388
H. Subsequent Legislation. .. 389
II. Jurisdiction and Standard of Review ... 389
A. Standard of Review Under the Administrative Procedure Act 389
B. Standard of Review Considerations Under Section 202(h) ... 390
*3801. “Determine whether any such rules are necessary in the public interest.”. rH 05 CO
2. “Repeal or modify any regulations it determines to be no longer in the public interest.” ^ 05 CO
C. Conclusion. LO 05 CO
III. Mootness and the National Television Ownership Rule. .395
IV. Cross-Ownership Rules . CO CD -q
A. Regulatory Background and the 2002 Biennial Review CO CO -3
B. The Commission’s decision not to retain a ban on newspaper/broadcast cross-ownership is justified under § 202(h) and is supported by record evidence. 00 05 CO
1. Newspaper/broadcast combinations can promote localism. 00 05 CO
2. A blanket prohibition on newspaper/broadcast combinations is not necessary to protect diversity. 399
C. The Commission’s decision to retain some limits on common ownership of different-type media outlets was constitutional and did not violate § 202(h). 400
1. Continuing to regulate cross-media ownership is in the public interest. 400
2. Continuing to regulate cross-media ownership does not violate the Fifth Amendment. 401
3. Continuing to regulate cross-media ownership does not violate the First Amendment. 401
D. The Commission did not provide reasoned analysis to support the specific Cross-Media Limits that it chose. 402
1. Overview of the Commission’s Diversity Index Methodology-403
2. The Commission did not justify its choice and weight of specific media outlets. 404
3. The Commission did not justify its assumption of equal market shares. OO O
4. The Commission did not rationally derive its Cross-Media Limits from the Diversity Index results. 05 O
5. The Commission should provide better notice on remand. rH rH
V. Local Television Ownership Rule . CM rH
A. Regulatory Background and the 2002 Biennial Review. CO rH
B. We uphold several threshold challenges to the Commission’s overall regulatory approach.
1. Limiting local television station ownership is not duplicative of antitrust regulation.
2. Media other than broadcast television may contribute to viewpoint diversity in local markets.
3. Consolidation can improve local programming. H-i CJT
4. The Commission adequately noticed its decision to allow triopolies. CD rH
C. We uphold the Commission’s decision to retain the top-four restriction. CD rH
D. We remand the specific numerical limits for the Commission’s further consideration.
E. We remand the Commission’s repeal of the Failed Station Solicitation Rule. O oq
VI. Local Radio Ownership Rule. to H
A. Regulatory Background and the 2002 Biennial Review. CO H
B. We uphold the Commission’s new definition of local markets. CO CO
1. The Commission justified using Arbitron Metro markets .... to CO
2. The Commission justified including noncommercial stations . to Ü1
C. We uphold the Commission’s transfer restriction. CO Oí
*3811. Transfer restriction is “in the public interest.”.427
2. Transfer restriction is reasoned decisionmaking.427
3. Transfer restriction is constitutional.428
D. We affirm the attribution of Joint Sales Agreements.429
1. Attribution of JSAs is “necessary in the public interest.”.429
2. Attribution of JSAs is reasoned decisionmaking. 429
3. Attribution of JSAs is constitutional..430
E. We remand the numerical limits to the Commission for further justification.430
1. The Commission’s numerical limits approach is rational and in the public interest .431
2. The Commission did not support its decision to retain the existing numerical limits with reasoned analysis.432
a. The Commission did not sufficiently justified “five equal-sized competitors” as the right benchmark.432
b. The Commission did not sufficiently justified that the existing numerical limits actually ensure that markets will have five equal-sized competitors.433
c. The Commission did not support its decision to retain the AM7FM subcaps.434
VII. Conclusion.435
In these consolidated appeals we consider revisions by the Federal Communications Commission to its regulations governing broadcast media ownership that the Commission promulgated following its 2002 biennial review. On July 2, 2003, the Commission announced a comprehensive overhaul of its broadcast media ownership rules. It increased the number of television stations a single entity may own, both locally and nationally; revised various provisions of the regulations governing common ownership of radio stations in the same community; and replaced two existing rules limiting common ownership among newspapers and broadcast stations (the newspaper/broadcast cross-ownership rule and the radio/television cross-ownership rule) with a single set of “Cross-Media Limits.” See Report and Order and Notice of Proposed Rulemaking, 18 F.C.C.R. 13,620, 2003 WL 21511828 (2003) (the “Order”).
Several public interest and consumer advocacy groups (collectively, the “Citizen Petitioners”)1 petitioned for judicial review of the Order in various courts of appeals, contending that its deregulatory provisions contravened the Commission’s statutory mandates as well as the Administrative Procedure Act, 5 U.S.C. §§ 551 et seq. (the “APA”). Associations of networks, broad*382casters, and newspaper owners2 also challenged the Order, arguing that pro-regulatory revisions as well as the absence of further deregulation violate the Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56 (1996) (the “1996 Act”), the APA, and the United States Constitution. The Judicial Panel on Multidistriet Litigation, acting pursuant to the random selection procedures of 28 U.S.C. § 2112(a), consolidated the petitions in this Court. On September 3, 2003, we stayed implementation of the rules pending our review.
For the reasons stated below, we affirm the power of the Commission to regulate media ownership. In doing so, we reject the contention that the Constitution or § 202(h) of the 1996 Act somehow provides rigid limits on the Commission’s ability to regulate in the public- interest. But we must remand certain aspects of the Commission’s Order that are not adequately supported by the record. Most importantly, the Commission has not sufficiently justified its particular chosen numerical limits for local television ownership, local radio ownership, and cross-ownership of media within local markets. Accordingly, we partially remand the Order for the Commission’s additional justification or modification, and we partially affirm the Order. The stay will continue pending our review of the Commission’s action on remand.3
I. Background
A. The 1934 Communications Act and Early Broadcast Ownership Regulation
In 1934 Congress authorized the Commission to grant licenses for private parties’ exclusive use of broadcast frequencies. Recognizing that the finite radio frequency spectrum inherently limits the *383number of broadcast stations that can operate without interfering with one another, Congress required that broadcast licensees serve the public interest, convenience, and necessity. Communications Act of 1934, 47 U.S.C. § 309(a); see also id. §§ 307(a), 310(d), 312.
“In setting its licensing policies, the Commission has long acted on the theory that diversification of mass media ownership serves the public interest by promoting diversity of program and service viewpoints, as well as by preventing undue concentration of economic power.” FCC v. Nat’l Citizens Comm. for Broad., 436 U.S. 775, 780, 98 S.Ct. 2096, 56 L.Ed.2d 697 (1978) (“NCCB”). The Commission’s early regulations reflected its presumption that a single entity holding more than one broadcast license in the same community contravened public interest. See Genesee Radio Corp., 5 F.C.C. 183, 186-87 (1938). In 1941, the Commission announced that it would not license more than one station in the same area to a single network organization. See Nat’l Broad. Co. v. United States, 319 U.S. 190, 206-08, 224-27, 63 S.Ct. 997, 87 L.Ed. 1344 (1943) (upholding the rule). At the same time, the Commission prohibited common ownership of stations within the same broadcast service (AM radio, FM radio, and television) in the same community. See Rules Governing Standard and High Frequency Broadcast Stations, 5 Fed. Reg. 2382, 2384 (June 26, 1940) (FM radio); Rules Governing Standard and High Frequency Broadcast Stations, 6 Fed. Reg. 2282, 2284-85 (May 6, 1941) (television); Rules Governing Standard and High Frequency Broadcast Stations, 8 Fed. Reg. 16065 (Nov. 27, 1943) (AM radio). Regulations limiting an entity to the common ownership of seven AM radio stations, seven FM radio stations, and seven television stations survived judicial scrutiny in 1956. See United States v. Storer Broad. Co., 351 U.S. 192, 76 S.Ct. 763, 100 L.Ed. 1081 (1956). In the 1970s the Commission adopted its first cross-ownership bans, which prohibited, on a prospective basis, the common ownership of television and radio stations serving the same market, as well as combinations of radio or broadcast stations with a daily newspaper in the same community. Amendment of Sections 73.35, 73.24,0 and 73.636 of the Commission Rules Relating to Multiple Ownership of Standard, FM and Television Broadcast Stations, 22 F.C.C.2d 306, ¶ 5, 1970 WL 18044 (1970); Amendment of Sections 73.34, 73.240 and 73.636 of the Commission’s Rules Relating to Multiple Ownership of Standard, FM, and Television Broadcast Stations, 50 F.C.C.2d 1046, 1975 WL 30457 (1975). The Supreme Court upheld the newspaper/broadcast cross-ownership ban as a “reasonable means of promoting the public interest in diversified mass communications.” NCCB, 436 U.S. at 802, 98 S.Ct. 2096.
B. Deregulation Initiatives
The 1980s saw a deregulatory trend for media ownership. The Commission raised its national ownership limits to permit common ownership of 12 stations in each broadcast service (though still prohibiting station combinations that would reach more than 25% of the national audience). Amendment of Section 73.3555 (formerly 73.35, 73.240, and 73.636) of the Commission’s Rules Relating to Multiple Ownership of AM, FM, and Television Broadcast Stations, 100 F.C.C.2d 74 ¶¶38, 39, 1985 WL 260060 (1985). The Commission also determined that UHF television stations should not be deemed to have the same audience reach as VHF stations,4 due *384to “the inherent physical limitations of [the UHF] medium,” and therefore applied a 50% “discount” to UHF audiences as counted under the national audience limitation. Id. ¶¶ 42-44. In other words, UHF stations count only half of their audiences in determining compliance with the national television ownership rule.
In 1989 the Commission eased its “one to a market” radio/television cross-ownership rules by allowing waiver requests for radio/television cross-ownership in the 25 largest television markets. The Commission stated that it would look favorably upon requests for waivers where there would be 30 independently owned broadcast “voices” remaining in the market after consolidation. Amendment of Section 73.3555 of the Commission’s Rules, the Broadcast Multiple Ownership Rules, 4 F.C.C.R. 1741, ¶ 1, 1989 WL 510875 (1989). In 1992, the Commission relaxed local and national radio ownership restrictions and adopted a tiered approach to radio concentration that allowed a single entity to own more radio stations in the largest markets (up to three AM and three FM stations, subject to a local audience reach limitation of 25% and a national cap of 30 AM stations and 30 FM stations) and fewer in the smallest markets. Revision of Radio Rules and Policies, 7 F.C.C.R. 6387, ¶ 27, 1992 WL 690638 (1992).
C. The Telecommunications Act of 1996
In 1996 Congress overhauled the Communications Act by enacting the Telecommunications Act of 1996. The 1996 Act contemplated a “pro-competitive, deregula-tory national policy framework designed to accelerate rapidly private sector development of advanced telecommunications and information technologies and services to all Americans by opening all telecommunications markets to competition.” S.Rep. No. 104-230, at 1-2 (1996). The 1996 Act eliminated all limits on national radio ownership and raised the national television audience reach cap from 25% to 35%. 1996 Act §§ 202(a), (c)(1)(B), 110 Stat. at 110-11. Congress also eased local radio ownership limits, establishing a four-tier sliding scale limit of numerical caps that allowed for as many as eight co-owned radio stations in the largest markets. Id. § 202(b)(1), 110 Stat. at 110.
The 1996 Act did not contain a new local television rule, but it directed the Commission to “conduct a rulemaking proceeding to determine whether to retain, modify, or eliminate” its existing local television ownership limitations. Id. § 202(c)(2), 110 Stat. at 111. It also expanded the applicability of the one-to-a-market radio/television cross-ownership restriction waiver to the fifty largest markets. Id. § 202(d), 110 Stat. at 111.
Finally, the 1996 Act instructed the Commission to review biennially its broadcast ownership rules “to determine whether any of such rules are necessary in the public interest as the result of competition.” Id. § 202(h), 110 Stat. at 111-12. Section 202(h) also required the Commission to “repeal or modify any regulation it determines to be no longer in the public interest.” Id.
D. Regulatory Review Since 1996
In 1999 the Commission responded to Congress’s directive under § 202(c) of the 1996 Act to review its local television rule and announced that it would relax its pro*385hibition on the common ownership of television stations with overlapping signals. The Commission’s new rule would allow two commonly owned television stations (a television station “duopoly”) in the same Designated Market Area (“DMA” or “market”) as long as (1) neither station was ranked among the four largest (“top-four”) stations in the market and (2) eight independently owned stations remained in the market post-merger. Review of the Commission’s Regulations Governing Television Broadcasting, 14 F.C.C.R. 12,903, ¶ 8, 1999 WL 591820 (1999) (“1999 Television Rule Review”). In the same rulemaking, the Commission also relaxed the one-to-a-market radio/television cross-ownership restriction and allowed radio/television station combinations to exist within three-tiered limits that depend on the size of the market. Id. ¶ 9.5
Meanwhile, in 1998 the Commission began the first biennial review of its broadcast ownership regulations as required under § 202(h). In 2000 it announced that it would retain the national television ownership rule (the 35% limit provided in the 1996 Act) and its cable/broadcast cross-ownership rule after determining that both rules remained “necessary in the public interest.” 1998 Biennial Regulatory Review — Review of the Commission’s Broadcast Ownership Rules and Other Rules Adopted Pursuant to Section 202 of the Telecommunications Act of 1996, 15 F.C.C.R. 11,058, ¶ 4, 2000 WL 791562 (2000) (“1998 Biennial Regulatory Review”)6 On appeal, however, the United States Court of Appeals for the D.C. Circuit held that the Commission had not sufficiently explained its reasons for retaining either of these rules. Fox Television Stations v. FCC, 280 F.3d 1027, 1043-44, 1051-52 (D.C.Cir.2002) (“Fox I”), modified on reh’g, 293 F.3d 537 (D.C.Cir.2002) (“Fox II”). As for the national television ownership rule, the Court determined that the agency had taken a “wait and see” approach in evaluating its competition and diversity effects, which was impermissible in light of § 202(h)’s mandate to repeal or modify rules found no longer necessary in the public interest. Id. at 1042. It remanded the rule to the Commission for additional justification. Id. at 1049. The Court vacated the cable/broadcast cross-ownership rule, however, finding that the Commission’s decision to retain it was arbitrary and capricious and contrary to § 202(h). Id. at 1053.
A few months later, the same Court reviewed the local television multiple-ownership rule. Sinclair Broad. Group, Inc. v. FCC, 284 F.3d 148 (D.C.Cir.2002). The petitioner challenged the “eight independent voices” exception, contending that it lacked foundation or connection to the Commission’s goal of promoting diversity in local markets. Id. at 152. The Court determined that the Commission had ade*386quately justified its decision to retain the local television rule under the APA and § 202(h), but that it had not provided a rational basis for the exclusion of non-broadcast media from the eight voices exception. Id. at 165. It remanded the rule for the Commission’s further justification. Id. at 169.
E. The Commission’s 2003 Report and Order
In September 2002 a Notice of Proposed Rulemaking (the “Notice”) announced that the Commission would review four of its broadcast ownership rules pursuant to § 202(h): the 35% national audience reach limit remanded in Fox; the local television rule remanded in Sinclair; the radio/television cross-ownership rule; and the dual network rule. 2002 Biennial Regulatory Review Notice of Proposed Rulemaking, 17 F.C.C.R. 18,503, ¶ 6, 2002 WL 31108252 (2002). The Notice also advised that the Commission was incorporating into its biennial review pending proceedings on two additional rules: its rule limiting radio station ownership in local markets and its rule prohibiting newspaper/broadcast cross-ownership. Id. ¶ 7.
The Commission established a Media Ownership Working Group (“MOWG”), which commissioned twelve studies ranging from consumer surveys to economic analyses of media markets. These reports were released for public comment in October 2002. Interested parties filed thousands of pages of comments, consisting of legal, social, and economic analyses, empirical and anecdotal evidence, and industry and consumer data to respond to the issues identified in the Commission’s Notice. Notably, nearly two million people weighed in by letters, postcards, e-mails, and petitions to oppose further relaxation of the rules. Statement of Commissioner Jonathan S. Adelstein, Dissenting, 18 F.C.C.R. 13,974, 13,977, 2003 WL 21251887 (July 2, 2003). The Commission also heard public comment at a February 2003 “field hearing” in Richmond, Virginia.7
On June 2, 2003, the Commission adopted the Order modifying its ownership rules to provide a “new, comprehensive framework for broadcast ownership regulation” by a vote of 3-2.8 Order ¶ 3. The Order was released on July 2, 2003.
F. The Order’s Modification of Broadcast Media Ownership Rules
After reaffirming the Commission’s three traditional policy objectives in promoting the public interest — competition, diversity, and localism — Order ¶ 8, the Commission considered whether each of the six rules remained in the public interest and proposed modifications where it believed necessary. With respect to each of the rules, the Commission determined as follows.
1. Local Television Ownership
The existing local television ownership rule allowed television station duopolies, so long as at least one of the stations was not ranked among the market’s four largest stations and so long as at least eight independently owned and operated full-power television stations would remain in the market post-merger. 47 C.F.R. § 73.3555(b). The Order modified this rule to permit television station triopolies *387in markets with 18 or more television stations and television station duopolies in markets with 17 or fewer television stations. Order ¶ 134. These limits are subject to the restriction (effectively a ban subject to a waiver provision) that a single firm may not own more than one top-four station in a market. This restriction forecloses common station ownership in markets with fewer than five television stations. Id. ¶ 186. The existing rule effectively precludes duopolies in most markets; only the largest 70 markets of the nation’s 210 DMAs could comply with the “eight voices” test. Adelstein Dissenting, 18 F.C.C.R. at 13,998. The new rule would allow triopolies in the nine largest DMAs, which represent 25.2% of the population. Duopolies could exist under the new rule in the largest 162 markets, representing 95.4% of the nation’s population. Id. at 13,997-98.
2. Local Radio Ownership
Although the Order retained the existing numerical limits on radio ownership that Congress established in § 202(b) of the 1996 Act,9 it modified other aspects of the rule. First, it changed the method for determining radio markets by replacing the “contour-overlap” method (described in detail in Part VLB infra) with the geography-based market delineations created by Arbitron, a company that generates market data for radio advertisers. Order ¶ 239. Additionally, the Commission would now include noncommercial stations in the station count for each market. Id. The Commission grandfathered any existing radio station combinations rendered noncompli-ant under the newly defined markets, id. ¶ 484, but generally restricted transfer of these combinations, id. ¶487. The Commission also changed the local radio ownership rule by deciding to attribute “Joint Sales Agreements” (agreements under which a licensee sells advertising time on its station to a broker station for a fee) toward the brokering entity’s numerical limit. Id. ¶ 239.
3 & 4. Newspaper/Broadcast and Radio/Television Cross-Ownership
The Commission has prohibited common ownership of a full-service television broadcast station and a daily public newspaper in the same community since 1975. Amendment of Sections 73.35, 73.240, and 73.636 of the Commission’s Rules Relating to Multiple Ownership of Standard, FM, and Television Broadcast Stations, 50 F.C.C.2d 1046, 1975 WL 30457 (1975). Additionally, the Commission regulates the number of television and radio stations that may be commonly owned with limits that vary with the size of the market. 47 C.F.R. § 73.3555(c).
The Order announced the Commission’s decision to repeal both cross-ownership rules (television/newspaper, radio/television) and replace them with a single set of Cross-Media Limits. The Commission determined that neither eross-ownOTship pro*388hibition remained necessary in the public interest to ensure competition, diversity, or localism. Order ¶¶ 330, 371. The new Cross-Media Limits prohibit newspaper/broadcast combinations and radio/television combinations in the smallest DMAs, ie., those with three or fewer full-power commercial or noncommercial television stations. Id. ¶454. In contrast, in the largest markets — those with more than eight television stations — common ownership among newspapers and broadcast stations is unrestricted. Id. ¶ 473. In medium-sized markets — those with between four and eight television stations — one entity may own a newspaper and either (a) one television station and up to 50% of the radio stations that may be commonly owned in that market under the local radio rule or (b) up to 100% of the radio stations allowed under the local radio rule. Id. ¶ 466.
In structuring its Cross-Media Limits, the Commission drew upon a methodological tool named the “Diversity Index,” which the Commission developed as a measure of viewpoint diversity in local markets to identify those “at-risk” markets where consolidation would have a deleterious effect. Id. ¶¶ 391, 442. The Diversity Index, explained more fully in Part IV.D.l below, is a highly modified version of the formula for measuring market concentration — the Herfindahl-Hirsehman Index— applied by the Department of Justice and Federal Trade Commission to analyze mergers. Id. ¶ 428.
5. National Television Ownership
The national television ownership rule prohibits entities from owning television stations that in the aggregate reach a certain percentage of our country’s households. 47 C.F.R. § 73.3555(e)(1). Section 202(c) of the 1996 Act directed the Commission to delete the then-existing twelve-station cap and raise the audience reach limit from 25% to 35%. After the D.C. Circuit Court remanded the Commission’s decision in the 1998 biennial review to retain the limit at 35%, Fox I, 280 F.3d at 1049, the Commission decided to increase the audience reach limit to 45%. Order ¶499. The Commission also declined to repeal or modify its existing 50% discount for UHF stations’ audiences as counted toward the audience reach limit. Id. ¶ 500.
6. Dual Network Rule
Under the dual network rule, a television station may affiliate with more than one network except that it may not affiliate with more than one of the four largest networks, ABC, CBS, Fox, and NBC. 47 C.F.R. § 73.658(g). The rule effectively permits common ownership of networks to the exclusion of the top four. Order ¶ 592. The Commission determined that the dual network rule remained necessary in the public interest, and thus did not repeal or modify it. Id.10
G. Procedural History of the Current Appeals
Within days of the publication of the Order, several organizations filed petitions for review of the Commission’s revised rules in various courts of appeals, some contending that the Commission had gone too far in revising the rules, and others asserting that the Commission had not gone far enough. Some of these organizations, including the Prometheus Radio Project, filed their petitions in this Court. Under 28 U.S.C. § 2112(a), petitions to review administrative orders filed in different circuit courts within the first ten days of the appeal period trigger a lottery conducted by the Judicial Panel on Multidis-*389trict Litigation. On August 19, 2003, the Panel announced that our Court had been selected in the lottery and consolidated the appeals here. We entered a stay of the effective date of the proposed rules after a hearing on September 3, 2003. Prometheus Radio Project v. FCC, No. 03-3388, 2003 WL 22052896 (3d Cir. Sept. 3, 2003). We then denied the Deregulatory Petitioners’ motion, joined by the Commission, to transfer venue to the D.C. Circuit Court on September 16, 2003. Prometheus Radio Project v. FCC, No. 03-3388 (3d Cir. Sept. 16, 2003) (order denying motion to transfer). After pushing back briefing and oral argument at the request of the parties, on February 11, 2004, we heard approximately eight hours of oral argument addressing the merits of Petitioners’ claims.
H. Subsequent Legislation
In January 2004, while the petitions to review the Order were pending in this Court, Congress amended the 1996 Act by increasing from 35% to 39% the national television ownership rule’s audience reach cap in § 202(c). Consolidated Appropriations Act, 2004, Pub. L. No. 108-199, § 629, 118 Stat. 3, 99 (2004). The legislation also amended § 202(h) in two ways: (1) making the Commission’s biennial review obligation quadrennial; and (2) insulating from § 202(h) review “rules relating to the 39 percent national audience reach limitation.” 118 Stat. at 100. Prior to oral argument, Petitioners filed letter briefs addressing' the 'effect of these amendments on their challenges to the Order.
II. Jurisdiction and Standard of Review
This is an appeal of an agency decision under the Communications Act of 1934, 47 U.S.C. §§ 151 et seq. Our jurisdiction is based on 47 U.S.C. § 402(a) and 28 U.S.C. § 2342(1). Our standard of review is governed by the APA and the 1996 Act provision authorizing the Commission’s periodic regulatory review.11
A. Standard of Review Under the Administrative Procedure Act
Our standard of review in the agency rulemaking context is governed first by the judicial review provision of the APA, 5 U.S.C. § 706. Under it, we “hold unlawful or set aside agency action, findings, and conclusions” that are found to be “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law ... [or] unsupported by substantial evidence.” Id. § 706(2)(a), (e); N.J. Coalition for Fair Broad. v. FCC, 574 F.2d 1119, 1125 (3d Cir.1978).
The scope of review under the “arbitrary and capricious” standard is “narrow, and a court is not to substitute its judgment for that of the agency.” Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983) (“State Farm”). Nevertheless, we must ensure that, in reaching its decision, the agency *390examined the relevant data and articulated a satisfactory explanation for its action, including a “rational connection between the facts found and the choice made.” Id. (quoting Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 9 L.Ed.2d 207 (1962)). Normally, we may find an agency rule is arbitrary and capricious where
the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. The reviewing court should not attempt itself to make up for such deficiencies; we may not supply a reasoned basis for the agency’s action that the agency itself has not given.
Id. (citing SEC v. Chenery Corp., 332 U.S. 194, 196, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947)); see also Robert Wood Johnson Univ. Hosp. v. Thompson, 297 F.3d 273, 280 (3d Cir.2002). Put another way, we reverse an agency’s decision when it “is not supported by substantial evidence, or the agency has made a clear error in judgment.” AT&T Corp. v. FCC, 220 F.3d 607, 616 (D.C.Cir.2000) (citing Kisser v. Cisneros, 14 F.3d 615, 619 (D.C.Cir.1994)).
We will, however, “uphold a decision of less than ideal clarity if the agency’s path may reasonably be discerned.” State Farm, 463 U.S. at 43, 103 S.Ct. 2856 (quoting Bowman Transp., Inc. v. Arkansas-Best Freight Sys., Inc., 419 U.S. 281, 286, 95 S.Ct. 438, 42 L.Ed.2d 447 (1974)). But an agency that departs from its “former views” is “obligated to supply a reasoned analysis for the change beyond that which may be required when an agency does not act in the first instance” in order to survive judicial scrutiny for compliance with the APA. State Farm, 463 U.S. at 41-42, 103 S.Ct. 2856.
Finally, the traditional APA standard of review is even more deferential “where the issues involve ‘elusive’ and ‘not easily defined’ areas such as programming diversity in broadcasting.” Sinclair, 284 F.3d at 159. Yet even when an administrative order involves policy determinations on such elusive goals, a “rationality” standard is appropriate. See NCCB, 436 U.S. at 796-97, 98 S.Ct. 2096 (finding that the Commission acted rationally in determining that diversification of ownership would enhance the possibility of increasing diverse viewpoints). Additionally, when an agency has engaged in line-drawing determinations and our review is necessarily deferential to agency expertise, see AT&T Corp., 220 F.3d at 627, its decisions may not be “patently unreasonable” or run counter to the evidence before the agency. Sinclair, 284 F.3d at 162.
B. Standard of Review Considerations Under Section 202(h)
The Order was promulgated as part of the periodic review requirements of § 202(h) of the 1996 Act. Consequently, our review standard is informed by that provision, which, at the time of the Order’s release, read:
(h) Further Commission Review. The Commission shall review its rules adopted pursuant to this section and all of its ownership rules biennially12 as part of its regulatory reform review under section 11 of the Communications Act of *3911934 and shall determine whether any of such rules are necessary in the public interest as the result of competition. The Commission shall repeal or modify any regulation it determines to be no longer in the public interest.
110 Stat. 111-12. Section 11 of the Communications Act, to which § 202(h) refers, was also added by the 1996 Act to ensure that the Commission review periodically its regulations governing telecommunications services to “determine whether any such regulation is no longer necessary in the public interest as a result of meaningful economic competition between providers of such service” and “repeal or modify any regulation it determines to be no longer necessary in the public interest.” 47 U.S.C. § 161.
The text and legislative history of the 1996 Act indicate that Congress intended periodic reviews to operate as an “ongoing mechanism to ensure that the Commission’s regulatory framework would keep pace with the competitive changes in the marketplace” resulting from that Act’s relaxation of the Commission’s regulations, including the broadcast media ownership regulations. 2002 Biennial Regulatory Review, 18 F.C.C.R. 4726, ¶¶ 16, 17 (2003) (citing preamble to the 1996 Act; H.R. Conf. Rep. No. 104-458 (1996)). Put another way, the periodic review provisions require the Commission to “monitor the effect of ... competition ... and make appropriate adjustments” to its regulations. Id. ¶ 5.
As noted, the first sentence of § 202(h) requires the Commission to “determine” whether media concentration rules are “necessary in the public interest as the result of competition.” The second sentence contains a separate instruction to the Commission: to “repeal or modify” those rules “no longer in the public interest.” 110 Stat. 111-12. We analyze each of these instructions in turn.
1. “Determine whether any such rules are necessary in the public interest.”
Recognizing that competitive changes in the media marketplace could obviate the public necessity for some of the Commission’s ownership rules, the first instruction requires the Commission to take a fresh look at its regulations periodically in order to ensure that they remain “necessary in the public interest.” This raises the question of what is “necessary.”
In the context of § 11 of the Communications Act, 47 U.S.C. § 161-which, like § 202(h), requires the Commission periodically to review its telecommunications regulations and determine whether they "remain necessary in the public interest"the Commission has interpreted "necessary" to mean "useful," "convenient" or "appropriate" rather than "required" or "indispensable." Setting out its rationale for this interpretation in the 2002 Biennial Regulatory Review, 18 F.C.C.R. 4726, ¶11 14-22, 2003 WL 1192543 (2003), the Commission determined that the 1996 Act's legislative history indicated that Congress meant "no longer necessary" to mean "no longer in the public interest" and "no longer meaningful." 18 F.C.C.R. 4726, 1117 (citing H.R. Conf. Rep. No. 104-458, at 185 (1996)).
Next, the Commission found that an "indispensable" construction of "necessary" as to § 11 would be unreasonably inconsistent with the Communications Act's grant of general rulemaking authority to the Commission. Id. at 1118 n. 31. Under 47 U.S.C. § 201(b) the Commission is authorized to "prescribe such rules and regTlla-tions [regarding services and charges of communications common carriers] as may be necessary in the public interest to carry *392out the provisions of this Act.” In AT&T Corp. v. Iowa Utilities Board, the Supreme Court interpreted this provision as a grant of “general rulemaking authority.” 525 U.S. 366, 374, 119 S.Ct. 721, 142 L.Ed.2d 835 (1999). Characterizing this interpretation “not as a limitation on the Commission’s authority, but a confirmation of it,” the Commission concluded that the standard of review applicable to its rule-making authority under § 201 is a “plain public interest” standard. 18 F.C.C.R. 4276, ¶¶ 18 n. 31, 22 (citing 525 U.S. at 374, 378, 119 S.Ct. 721). The Commission reasoned that the same standard must also apply to the review process required under § 11 in order to avoid absurd results. If the rulemaking and review standards were different, the Commission could promulgate any rule that is useful, but then, at the next periodic review, would have to revoke any of those rules that do not also meet a higher standard of “indispensable.” Id. ¶ 18 & n. 33. Under such a system, periodic review would either be inefficient or irrelevant, as the Commission could effectively sidestep the more stringent review standard by subsequently reissuing any “useful” rule that it had to repeal for failing to be “indispensable.” Id. ¶ 18.
Lastly, the Commission rejected arguments that there is controlling judicial precedent for an “indispensable” construction of “necessary.” It acknowledged that the Supreme Court and the D.C. Circuit Court of Appeals have upheld such constructions, see id. ¶ 19 (citing Iowa Utils. Bd., 525 U.S. at 374, 378, 119 S.Ct. 721; GTE Serv. Corp. v. FCC, 205 F.3d 416 (D.C.Cir.2000)), but countered that these cases “simply demonstrate that terms such as ‘necessary’ ... must be read in their statutory context.” Id. Furthermore, the Commission found judicial support for its interpretation of “necessary” in the D.C. Circuit Court’s decision in Sinclair, 284 F.3d at 159, regarding the Commission’s periodic review of its local television ownership rule under § 202(h). The Commission noted that the Sinclair Court did not expressly adopt any particular definition of “necessary,” but, in affirming the Commission’s § 202(h) finding that the rule furthers diversity and is thus necessary in the public interest, id. at 160, it “did not articulate a new or higher public interest yardstick.” 18 F.C.C.R. 4726, ¶20.13 Nor did the D.C. Circuit Court’s decision in Fox foreclose its interpretation of “necessary,” the Commission determined, because on rehearing the Court deleted language in its initial decision that the Commission had applied too lax a standard in reviewing its broadcast media ownership rules under § 202(h).14 Id. ¶ 14 (citing Fox II, 293 F.3d at 540).
For these reasons, the Commission determined that § ll’s requirement that it review its telecommunications regulations to determine whether they remain “necessary in the public interest” does not require it to employ a more stringent standard than “plain public interest” found in other parts of the Communications Act. Id. ¶¶ 18, 22 (citing 47 U.S.C. § 201(b) as an example).
*393Recently, the D.C. Circuit Court upheld, in the context of § 11, the Commission’s interpretation of “necessary” contained in the 2002 Biennial Regulatory Review. Cellco P’ship v. FCC, 357 F.3d 88 (D.C.Cir.2004). Recognizing that “necessary” is a “chameleon-like” word whose “meaning ... may be influenced by its context,” the Célico Court determined that it would uphold any reasonable interpretation that did not contravene the express provisions of the Communications Act. Id. at 94, 96 (citing Chevron U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984)). It went on to determine that the Commission’s interpretation is both reasonable and consistent with the Communications Act, endorsing the Commission’s view that “necessary” must mean the same thing in the periodic review context as in the rule-making context in order to avoid absurd results. Id. at 98.
Célico also acknowledged that the Commission’s interpretation of “necessary” is consistent with the many courts that have endorsed a “useful” or “appropriate” interpretation over an “essential” or “indispensable” one. Id. at 97 (citing, inter alia, NCCB, 436 U.S. at 795-96 98 S.Ct. 2096; McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 413, 4 L.Ed. 579 (1819); Cellular Telecomm. & Internet Ass’n v. FCC, 330 F.3d 502, 510 (D.C.Cir.2003) (specifically rejecting an “indispensable” connotation of “necessary” as used in the Communications Act’s enforcement forbearance provision, § 10(a))).
Finally, the Célico Court rejected suggestions that the Commission’s interpretation was inconsistent with its prior decisions in Sinclair and Fox. As noted above, Sinclair did not expressly adopt any particular definition of “necessary” and Fox I’s suggestion of a heightened standard was expressly retracted by Fox II, 293 F.3d at 540. Cellco limited Fox I’s statement that “necessary” implied a presumption in favor of modification or elimination of existing regulations, see 280 F.3d at 1048, to the context in which it was made: discussing whether vacating or remanding the national television ownership rule was the appropriate remedy. Cellco, 357 F.3d at 98. And while Sinclair apparently endorsed this language from Fox I, see 284 F.3d at 159, the Cellco Court characterized Sinclair as merely “piggybacking]” on Fox I without “adopting] a general presumption in favor of modification or elimination of regulations when considering a substantive challenge to the adequacy of the Commission’s determinations.” Cellco, 357 F.3d at 98. In sum, the D.C. Circuit Court determined that the definition of “necessary” was not constrained by either its Fox or Sinclair decision. It remained an open issue for the Commission to decide in the first instance, as it did when it released the 2002 Biennial Regulatory Review. Id.
For the same reasons proffered by the Commission and endorsed by the D.C. Circuit Court to reject the “indispensable” definition of “necessary” under § 11, we do so under § 202(h). Though § 11 and § 202(h) are separate statutory provisions, they are both periodic review provisions from the same statute. As evidence of their relatedness, § 202(h) imposes periodic review obligations “as part of its regulatory reform review under section 11 of this Communications Act of 1934,” 110 Stat. at 111.15 We see no reason to adopt a different definition of “necessary” under § 202(h) than under § 11. Moreover, interpreting § 202(h)’s first sentence to require the Commission to review its rules to *394determine whether they are indispensable in the public interest would lead to incongruous results when compared to the instruction in § 202(h)’s second sentence, which requires the Commission to “repeal or modify any regulation it determines to be no longer in the public interest.”16 For the “determine” instruction to be meaningful, “necessary” must embody the same “plain public interest” standard that Congress set out in the “repeal or modify” instruction.17 Lastly, as explained by the Célico Court, the “convenient,” “useful,” or “helpful” definition of “necessary” is not foreclosed to the Commission by any judicial precedent, including Fox and Sinclair. So in interpreting the Commission’s obligation under § 202(h) to review its broadcast media ownership rules to determine whether they are “necessary in the public interest,” we adopt what the Commission termed “the plain public interest” standard under which “necessary” means “convenient,” “useful,” or “helpful,” not “essential” or “indispensable.”
2. “Repeal or modify any regulations it determines to be no longer in the public interest.”
Turning to the second instruction of § 202(h), the Commission is required to “repeal or modify” rules that are “no longer in the public interest.” Having concluded that the first instruction requires the Commission to determine whether any existing rule fails to satisfy the “plain public interest” standard, the relationship between the first and second instruction is evident. Under the second instruction, the Commission must repeal or modify the regulations that it has determined under the first instruction do not satisfy that same standard.
While we acknowledge that § 202(h) was enacted in the context of deregulatory amendments (the 1996 Act) to the Communications Act, see Fox I, 280 F.3d at 1033; Sinclair, 284 F.3d at 159, we do not accept that the “repeal or modify in the public interest” instruction must therefore operate only as a one-way ratchet, ie., the Commission can use the review process only to eliminate then-extant regulations. For starters, this ignores both “modify” and the requirement that the Commission act “in the public interest.” What if the Commission reasonably determines that the public interest calls for a more stringent regulation? Did Congress strip it of the power to implement that determina*395tion? The obvious answer is no, and it will continue to be so absent clear congressional direction otherwise.18
What, then, makes § 202(h) “deregula-tory”? It is this: Section 202(h) requires the Commission periodically to justify its existing regulations, an obligation it would not otherwise have. A regulation deemed useful when promulgated must remain so. If not, it must be vacated or modified.
Misguided by the Fox and Sinclair Courts’ “deregulatory presumption” characterization and lacking the benefit of Cell-co’s subsequent clarification, the Commission concluded that § 202(h) “appears to upend traditional administrative law principles” by not requiring it to justify affirmatively a rule’s repeal or modification. Order ¶ 11. This overstates the case. Rather than “upending” the reasoned analysis requirement that under the APA ordinarily applies to an agency’s decision to promulgate new regulations (or modify or repeal existing regulations), see State Farm, 463 U.S. at 43, 103 S.Ct. 2856, § 202(h) extends this requirement to the Commission’s decision to retain its existing regulations. This interpretation avoids a crabbed reading of the statute under which we would have to infer, without express language, that Congress intended to curtail the Commission’s rulemaking authority and to contravene “traditional administrative law principles.”19
C. Conclusion
Though our standard of review analysis is lengthy, it is in the end amenable to a straightforward summing-up: In a periodic review under § 202(h), the Commission is required to determine whether its then-extant rules remain useful in the public interest; if no longer useful, they must be repealed or modified.20 Yet no matter what the Commission decides to do to any particular rule-retain, repeal, or modify (whether to make more or less stringent)-it must do so in the public interest and support its decision with a reasoned analysis. We shall evaluate each aspect of the Commission's Order accordingly.
III. Mootness and the National Television Ownership Rule
The national television ownership rule caps the number of television stations *396that a single entity may own on a national basis. In the 1996 Act, Congress limited the number of commonly owned stations to those reaching no more than 35% of the national audience.1996 Act § 202(c)(1)(B), 110 Stat. 111. In its first biennial review, the Commission retained the 35% cap as “necessary in the public interest,” but the D.C. Circuit Court held that the Commission had not adequately justified this decision. Fox I, 280 F.3d at 1048. On remand and in connection with its 2002 biennial review proceeding, the Commission increased the cap from 35% to 45%. Order ¶ 583. The Commission also decided to retain its method of discounting by 50% the audiences of UHF stations toward the cap. Id. ¶ 586.
Subsequently, however, Congress enacted a new national television ownership cap. In its 2004 Consolidated Appropriations Act, it modified § 202(c)(1)(B) of the 1996 Act to provide that “[t]he Commission shall modify its rules for multiple ownership ... by increasing the national audience reach limitation for television stations to 39%.” See Pub. L. No. 108-199, § 629, 118 Stat. 3, 99 (2004). Because the Commission is under a statutory directive to modify the national television ownership cap to 39%, challenges to the Commission’s decision to raise the cap to 45% cap are moot.21 Cf. PLMRS Narrowband Corp. v. FCC, 182 F.3d 995, 1002 (D.C.Cir.1999) (challenges to Commission’s later-modified order are moot).
Although the 2004 Consolidated Appropriations Act did not expressly mention the UHF discount, challenges to the Commission’s decision to retain it are likewise moot. Congress instructed the Commission to “increase the national audience reach limitation for television stations to 39%.” 118 Stat. at 99. Since 1985 the Commission has defined “national audience reach” to mean “the total number of television households” reached by an entity’s stations, except that “UHF stations shall be attributed with 50 percent of the television households” reached. 47 C.F.R. § 73.3555(e)(2)(i); Multiple Ownership of AM, FM and Television Broadcast Stations, 50 Fed. Reg. 4666, 4676, 1985 WL 260060 (Feb. 1, 1985). We assume that when Congress uses an administratively defined term, it intended its words to have the defined meaning. See, e.g., Bragdon v. Abbott, 524 U.S. 624, 631, 118 S.Ct. 2196, 141 L.Ed.2d 540 (1998). Furthermore, because reducing or eliminating the discount for UHF station audiences would effectively raise the audience reach limit, we cannot entertain challenges to the Commission’s decision to retain the 50% UHF discount. Any relief we granted on these claims would undermine Congress’s specification of a precise 39% cap.
As additional evidence of the mootness of challenges to the UHF discount, we note that the 2004 Consolidated Appropriations Act also added a sentence to § 202(h): “This subsection does not apply to any rules relating to the 39% national audience limitation.” 118 Stat. at 100. *397The UHF discount is a rule “relating to” the national audience limitation. See 47 C.F.R. § 73.3555(e)(2) (providing for the UHF discount in a section qualified “for the purposes of this paragraph (e),” the national television ownership rule paragraph). Congress apparently intended to insulate the UHF discount from periodic review, a position that is consistent with our reading of the legislation as endorsing the almost 20-year-old regulatory definition of “national audience reach” that provides for the UHF discount.
Although we find that the UHF discount is insulated from this and future periodic review requirements, we do not intend our decision to foreclose the Commission’s consideration of its regulation defining the UHF discount in a rulemaking outside the context of Section 202(h). The Commission is now considering its authority going forward to modify or eliminate the UHF discount and recently accepted public comment on this issue. 69 Fed. Reg. 9216-17 (Feb. 27, 2004). Barring congressional intervention, see, e.g., S. 1264, 108th Cong. § 12 (2003) (proposing phase-out and 2008 sunset of the UHF), the Commission may decide, in the first instance, the scope of its authority to modify or eliminate the UHF discount outside the context of § 202(h).
IV. Cross-Ownership Rules
The Commission’s decision to repeal its newspaper/broadcast cross-ownership rules22 in favor of new Cross-Media Limits has been attacked on all fronts. Some petitioners support the repeal but argue that the Cross-Media Limits are too restrictive. Others challenge the repeal decision and argue that the new limits are too lenient. We conclude that the Commission’s decision to replace its cross-ownership rules with the Cross-Media Limits is not of itself constitutionally flawed and does not violate § 202(h). But we cannot uphold the Cross-Media Limits themselves because the Commission does not provide a reasoned analysis to support the limits that it chose.
A. Regulatory Background and the 2002 Biennial Review
Since the 1970s, the Commission has enforced two separate limits on the common ownership of different-type media outlets in local markets. One cross-ownership'rule prohibits the common ownership of a full-service television-broadcast station and a daily public newspaper in the same community. 47 C.F.R. § 73.3555(d). The other limits the number of television and radio stations to the following combinations: (1) in.markets where at least 20 independently owned media, voices would remain post-merger, two television stations and six radio stations or one television station and seven radio stations; (2) in markets where at least 10 independent voices would remain, two television stations and four radio stations; and (3) in other markets, two television stations (subject to the local television ownership rule) and one radio station. Id. § 73.3555(c).
The Commission’ considered both cross-ownership rules during its 2002 biennial review under § 202(h). In the Order, the Commission announced that because neither rule remained necessary in the public interest, it was repealing them and replacing them with a single set of Cross-Media Limits. The three-tiered Cross-Media Limits regulate common ownership depending on the size of the market: small (those with three or fewer full-power commercial or noncommercial television sta*398tions), mid-sized (between four and eight television stations), and large (more than eight television stations). In small markets, newspaper/broadcast combinations and radio/television combinations are prohibited. Order ¶ 454. In medium-sized markets, an entity may own a newspaper and either (a) one television station and up to 50% of the radio stations that may be commonly owned in that market under the local radio rule or (b) up to 100% of the radio stations allowed under the local rule. Id. ¶ 466. In large markets, cross-ownership is unrestricted. Id. ¶ 473.
B. The Commission’s decision not to retain a ban on newspaper/broadcast cross-ownership is justified under § 202(h) and is supported by record evidence.
The Commission determined that the rule prohibiting newspaper/broadcast cross-ownership was no longer necessary in the public interest for three primary reasons: (1) the ban is not necessary to promote competition in local markets because most advertisers do not view newspapers and television stations as close substitutes, Order ¶ 332; (2) the ban undermines localism by preventing efficient combinations that would allow for the production of high-quality local news, id. ¶ 343; and (3) there is not enough evidence to conclude that ownership influences viewpoint to warrant a blanket cross-ownership ban, thus making it unjustifiable on diversity grounds, id. at Ii 364, (and moreover, the presence of other media sources-such as the Internet and cable-compensate for the viewpoint diversity lost to consolidation, id. ¶ 365). The Citizen Petitioners object to the local-ism and diversity components of the Commission's rationale. We conclude differently, as reasoned analysis supports the Commission's determination that the blanket ban on newspaper/broadcast cross-ownership was no longer in the public interest. Part II.C supra.
1. Newspaper/broadcast combinations can promote localism.
The Commission measured the promotion of localism by considering “the selection of programming responsive to local needs and interests, and local news quantity and quality.” Order ¶ 78. Evidence that existing (grandfathered) newspaper-owned broadcast stations produced local news in higher quantity with better quality than other stations convinced the Commission that the ban on newspaper/broadcast combinations undermined its localism interest. The Commission principally relied on the findings of its MOWG study that newspaper-owned television stations provide almost 50% more local news and public affairs programming than other stations, an average of 21.9 hours per week. Id. ¶ 344 (citing Thomas C. Spavins et al., The Measurement of Local Television News and Public Affairs Programs (MOWG Study No. 7) at 3 (Sept.2002)). The Commission also found corresponding advantages in quality of local coverage provided by newspaper-owned stations, as shown by ratings (measuring consumer approval) and industry awards (measuring critical approval). Id. ¶ 344-45 (citing, among other things, findings by the Project for Excellence in Journalism that newspaper-owned stations “were more likely to do stories focusing on important community issues and to provide a wide mix of opinions, and they were less likely to do celebrity and human interest features”).
The Citizen Petitioners argue that the MOWG study was flawed because it examined all newspaper/broadcast station combinations, including “intermarket” combinations (entities that own a newspaper and broadcast stations in different cities), as *399well as “intramarket” combinations (entities that own a newspaper and a broadcast, station in the same city). But the Citizens Petitioners do not suggest that a study entirely focused on intramarket combinations would have different results. The six intramarket combinations that were included in the study (grandfathered exceptions to the cross-ownership ban) averaged more local news and public affairs programming as compared to the overall average (26 weekly hours compared to 21.9) and higher ratings for their 5:30 p.m. and 6:00 p.m. news programs (9.8 and 11 compared to 7.8 and 8.2). MOWG Study No. 7 app. A; Comments of Newspaper Association of America, MB Docket No. 02-277 at 15 (Jan. 2, 2003).
The Citizen Petitioners also protest the Commission’s reliance on anecdotal evidence of pro-localism combinations and its disregard of anecdotal evidence to the contrary. Significantly, however, the Commission used anecdotal evidence23 merely to illustrate its statistical findings — it did not rely on anecdote as the sole basis for its conclusions about localism. Moreover, the Commission properly discounted anecdotal evidence of a Canadian newspaper conglomerate’s detrimental effect on local-ism. Can West, which controls 30% of Canada’s daily newspaper circulation, requires its newspaper editors to publish editorials from headquarters, which it forbids local editorials to contradict. But the Commission explained that the Can West example shows the peril of national ownership and corporate centralization of media services, which is not relevant to the Commission’s regulations on local combinations. Order ¶ 352. In summary, the Citizen Petitioners’ arguments do not unsettle the Commission’s conclusion that the newspaper/broadcast cross-ownership ban undermined localism.
2. A blanket prohibition on newspaper/broadcast combinations is not necessary to protect diversity.
The Commission offered two rationales for its conclusion that a blanket prohibition on newspaper/broadcast combinations is no longer necessary to ensure diversity in local markets. First, it found that “[Commonly-owned newspapers and broadcast stations do not necessarily speak with a single, monolithic voice.” Id. ¶ 361. Given conflicting evidence in the record24 on whether ownership influences viewpoint, the Commission reasonably concluded that it did not have enough confidence in the proposition that commonly owned outlets *400have a uniform bias to warrant sustaining the cross-ownership ban. Id. ¶ 864.
Second, the Commission found that diverse viewpoints from other media sources in local markets (such as cable and the Internet) compensate for viewpoints lost to newspaper/broadcast consolidations. Id. ¶ 366. We agree record evidence suggests that cable and the Internet supplement the viewpoint diversity provided by broadcast and newspaper outlets in local markets.25 As discussed more fully below, we believe that the Commission gave too much weight to the Internet in deriving the Cross-Media Limits. But separate from the question of degree, we conclude that it was acceptable for the Commission to find that cable and the Internet contribute to viewpoint diversity.
C. The Commission’s decision to retain some limits on common ownership of different-type media outlets was constitutional and did not violate § 202(h).
The Deregulatory Petitioners support the Commission’s repeal of the newspaper/broadcast cross-ownership ban but object to its decision to retain any restriction on the common ownership of newspaper and broadcast media outlets. First they argue that any limits on newspaper/broadcast cross-ownership violate § 202(h), because, as the Commission acknowledges, the evidence suggests that the cross-ownership restrictions are not in the public interest. They also argue that the Cross-Media Limits violate the First and Fifth Amendments of the United States Constitution. We disagree on all counts.
1. Continuing to regulate cross-media ownership is in the public interest.
The Commission’s finding that a blanket prohibition on newspaper/broadcast cross-ownership is no longer in the public interest does not compel the conclusion that no regulation is necessary. See Order ¶ 364. As described above, the Commission found evidence to undermine the premise that ownership always influences viewpoint,26 but it did not find the opposite to be true. And while the Commission found that other media sources contributed to viewpoint diversity in local markets, it could not have found that the Internet and cable were complete substitutes for the viewpoints provided by newspapers and broadcast stations. See MOWG Study No. 3 (finding that the Internet and cable rank as sources' of local news, but they do not outrank newspapers and broadcast television). Given the Commission’s goal of balancing the public’s interests in competition, localism, and diversity, it reasonably concluded that repealing the cross-owner*401ship ban was necessary to promote competition and localism, while retaining some limits was necessary to ensure diversity.27
2. Continuing to regulate cross-media ownership does not violate the Fifth Amendment.
The Deregulatory Petitioners argue that the Cross-Media Limits violate the Equal Protection Clause of the Fifth Amendment because they single out newspaper owners for special restrictions that do not apply to other media outlets. This argument is foreclosed, however, by the Supreme Court’s decision in NCCB that endorsed the constitutionality of the 1975 newspaper/television cross-ownership ban. 436 U.S. at 801-02, 98 S.Ct. 2096. Rather than concluding that newspaper owners were singled out for different treatment, the Court determined that “the regulations treat newspaper owners in essentially the same fashion as other owners of the major media of mass communication” by imposing limits on them as well as owners of television and radio stations. Id. at 801, 98 S.Ct. 2096. We decline the Deregula-tory Petitioners’ invitation to disregard Supreme Court precedent because of changing times. Surely there are more media outlets today (such as cable, the Internet, and satellite broadcast) than there were in 1978 when NCCB was decided. But it cannot be assumed that these media outlets contribute significantly to viewpoint diversity as sources of local news and information. See Part IV.D.2 infra. Even if it could, it is the Supreme Court’s prerogative to change its own precedent. See State Oil Co. v. Khan, 522 U.S. 3, 20, 118 S.Ct. 275, 139 L.Ed.2d 199 (1997) (stating that only the Supreme Court can overrule its own precedent); Am. Civil Liberties Union of N.J. v. Black Horse Pike Reg’l Bd. of Educ., 84 F.3d 1471, 1484 (3d Cir.1996) (stating that we are obliged to follow Supreme Court precedent “until instructed otherwise by a majority of the Supreme Court”).
3. Continuing to regulate cross-media ownership does not violate the First Amendment.
The Deregulatory Petitioners argue that the Commission’s decision to retain restrictions on the common ownership of newspapers and broadcast stations contravenes the First Amendment because it limits the speech opportunities of newspaper owners and broadcast station owners, and hence limits the public’s access to information. Yet again their challenge is foreclosed by NCCB, where the Supreme Court affirmed the Commission’s authority, despite a First Amendment challenge, to regulate broadeasVnewspaper cross-ownership in the public interest. Due to the “physical scarcity” of the broadcast spectrum, the Court scrutinized the regulation to discern a rational basis. 436 U.S. at 799, 98 S.Ct. 2096. The Commission’s action, it held, was “a reasonable means of promoting the public interest in diversified mass communications.” Id.’ at 802, 98 S.Ct. 2096.
The Deregulatory Petitioners suggest, as they did in mounting their Fifth Amendment challenge, that the expansion of media outlets since NCCB’s day requires a rethinking of the scarcity rationale and the lower level of constitutional review it entails. Again we decline their invitation to disregard precedent, and we are not alone. See FCC v. League of Women Voters of Cal., 468 U.S. 364, 376 & n. 11,104 S.Ct. 3106, 82 L.Ed.2d 278 (1984) *402(upholding the scarcity rationale until Congress speaks to the issue); Fox I, 280 F.3d at 1046 (“First, contrary to the implication of the networks’ argument, this court is not in a position to reject the scarcity rationale even if we agree that it no longer makes sense. The Supreme Court has already heard the empirical case against that rationale and still ‘declined to question its continuing validity.’ ” (citing Turner v. FCC, 512 U.S. 622, 638, 114 S.Ct. 2445, 129 L.Ed.2d 497 (1994))).
Even were we not constrained by Supreme Court precedent, we would not accept the Deregulatory Petitioners' contention that the expansion of media outlets has rendered the broadcast spectrum less scarce. In NCCB, the Court referred to the "physical scarcity" of the spectrum-the fact that many more people would like access to it than can be accommodated. 436 U.S. at 799, 98 S.Ct. 2096. The abundance of non-broadcast media does not render the broadcast spectrum any less scarce. See, e.g., Ruggiero v. FCC, 278 F.3d 1323, 1325 (D.C.Cir.2002), rev'd en banc, 317 F.3d 239 (D.C.Cir.2003) (citing the Commission's statement that "[niow radio service is widely available throughout the country and very little spectrum remains available for new full-powered stations.").
In this context, we will apply a rational basis standard to the Commission’s restrictions on the common ownership between newspaper and broadcast stations, and uphold them if they are rationally related to a substantial government interest. See NCCB, 436 U.S. at 799-800, 98 S.Ct. 2096; see also Am. Family Ass’n v. FCC, 365 F.3d 1156, 1168 (D.C.Cir.2004). In NCCB, the Supreme Court endorsed a substantial government interest in promoting diversified mass communications. Id. at 795, 802, 98 S.Ct. 2096. The Supreme Court held that the Commission had “acted rationally in finding that diversification of ownership would enhance the possibility of achieving greater diversity of viewpoints.” Id. at 796, 98 S.Ct. 2096. Here, as in NCCB, the Commission justified its continued restrictions on common ownership of newspapers and broadcast stations as promoting the public interest in viewpoint diversity. Order ¶ 355. The Court has said that limiting common ownership is a reasonable means of promoting the public interest in viewpoint diversity. NCCB, 436 U.S. at 796, 436 U.S. 775. Therefore, applying NCCB, we hold that the Commission’s continued regulation of the common ownership of newspapers and broadcasters does not violate the First Amendment rights of either.
D. The Commission did not provide reasoned analysis to support the specific Cross-Media Limits that it chose.
The Commission concluded that cross-ownership limits were necessary in specific situations to guard against "an elevated risk of harm to the range and breadth of viewpoints that may be available to the public." Order ¶ 442. But recognizing that ownership limits impede the speech opportunities for both broadcasters and newspapers, the Commission endeavored to craft new limits "as narrowly as possible." Id. ¶ 441. In that vein, the Commission sought to identify "at risk" local markets-those with high levels of viewpoint concentration-where continued regulation was necessary. By focusing its regulation on those markets, the Commission hoped to avoid needlessly overregulating markets with already ample viewpoint diversity.
But for all of its efforts, the Commission's Cross-Media Limits employ several irrational assumptions and inconsistencies. We do not object in principle to the Corn-*403mission’s reliance on the Department of Justice and Federal Trade Commission’s antitrust formula, the Herfindahl-Hirsch-mann Index (“HHI”), as its starting point for measuring diversity in local markets. In converting the HHI to a measure for diversity in local markets, however, the Commission gave too much weight to the Internet as a media outlet, irrationally assigned outlets of the same media type equal market shares, and inconsistently derived the Cross-Media Limits from its Diversity Index results. For these reasons, detailed below, we remand for the Commission to justify or modify further its Cross-Media Limits.
1. Overview of the Commission’s Diversity Index Methodology
The Commission designed a methodology called the Diversity Index to identify “at risk” markets where limits on cross-media ownership should be retained. The Diversity Index was “inspired by” the HHI, which the Department of Justice and the Federal Trade Commission use to measure proposed mergers’ effect on competition in local markets. Id. ¶ 394. A market’s HHI score is the sum of market shares squared. A highly competitive market will have a lower HHI score than a concentrated one. For example, compare the HHI score of a market with 10 equal-sized competitors (10 [competitors] x 10 [each competitor’s market share percentage]2 = 1000) to the HHI score of a market with only five equal-sized competitors (5 x 202 = 2000).28 The Department of Justice and the Federal Trade Commission categorize markets with HHI scores above 1800 as “highly concentrated.” If a proposed merger would exceed that level of concentration, the agencies believe that it would be harmful to the competition in that market. At its core, the Diversity Index here uses the same sum of market share-squared formula.
First, the Commission selected which media outlets to include in its analysis of viewpoint diversity in local markets based on consumers’ reported preferences for sources of local news and information. The Commission determined that broadcast television, daily and weekly newspapers, radio, and Internet (via cable connection and DSL, dial-up, or other connections) are the relevant contributors to viewpoint diversity in local markets. See Neilsen Media Research, Consumer Survey on Media Usage (MOWG Study No. 8) (Sept. 2002). Based on the popularity of each media source, the Commission assigned each a relative weight: broadcast TV 33.8%, daily newspapers 20.2%, weekly newspapers 8.6%, radio 24.9%, Internet (cable) 2.3% and Internet (DSL, dial-up, or other connection) 10.2%. Order ¶¶ 412, 415, 427.
Next, the Commission selected many sample markets29 for which it would determine a Diversity Index score. In each of those markets, it counted the number of outlets within each media type and assigned each outlet an equal market share. (For example, the New York City market has 23 television stations, so each one was attributed an equal 4.3%30 market share.) *404The Commission calculated the ownership share by multiplying the number of outlets owned by an entity by the market share. (Univision owns three television stations in New York, so its ownership share was 3 x 4.3% = 13.0%.) Each ownership share was then given its relative weight by media type. (Univision’s 13.0% share was thus subject to the 33.8% multiplier for television stations). The Commission then squared all of the weighted ownership shares; their sum was the market’s total Diversity Index score. But in markets with cross-owned shares (outlets of different media types owned by the same entity) the entity’s weighted ownership shares were summed together before they were squared. (ABC owned one television station and four radio stations in New York, so the Commission added its weighted ownership share for the television station (4.3% x 33.8% = 1.45%) to its weighted ownership share for the radio stations (6.7% x 24.9% = 1.67%) for a combined ABC ownership share of 3.12%. The square of 3.12 was included in the summation). See generally Order app. C.
After the Commission calculated Diversity Index scores for markets of different sizes, it determined how those scores would change in several different consolidation “scenarios.” To illustrate, the Commission determined that markets with five television stations had an average Diversity Index score of 911. If a newspaper and television station combined in a market of that size, that score would increase by 223, to 1134. The other combination scenarios were: (1) one television station and all of the radio stations allowed to be commonly owned under the local radio rule; (2) one newspaper and all of the radio stations allowed to be commonly owned under the Iocal radio rule; (3) one newspaper, one television station, and half of the number of radio stations allowed to be commonly owned under the local radio rule; (4) two television stations; (5) one newspaper and two television stations; and (6) one newspaper, two television stations, and all of the radio stations allowed under the local radio rule. Order app. D.
Finding that all of the consolidation scenarios resulted in relatively high increases to the average Diversity Index scores for the smallest markets (those with three or fewer television stations), the Commission prohibited newspaper/television, newspaper/radio, and radio/television combinations in those markets. Order ¶¶ 456, 459, 460. In the large markets (nine or more stations), all of the consolidation scenarios resulted in acceptable increases to the average Diversity Index scores, so the Commission imposed no limits on cross-media ownership in those markets. In the midsized markets (between four and eight television stations), the Commission found that all of the scenarios, except the two involving a newspaper and television duopoly, should be allowed, based on their modest increases to- the average Diversity Index scores for those markets. Id. ¶ 466.
2. The Commission did not justify its choice and weight of specific media outlets.
Petitioners from both sides of the regulatory spectrum attack the Commission’s selection of media outlets for inclusion in the Diversity Index formula. The Citizen Petitioners argue that the Commission gave too much weight to the Internet at the expense of television and daily newspapers, thereby understating the level of concentration and overstating the level of di*405versity in a market. The Deregulatory Petitioners, on the other hand, argue that the Diversity Index understates the actual amount of diversity in a market by ignoring important media outlets, primarily cable television. As explained below, we affirm the Commission’s reasoned decision to discount cable. But we think that the same rationale also applies to the Internet. Therefore, its decision to count the Internet as a source of viewpoint diversity, while discounting cable, was not rational.
The Commission properly excluded cable because of serious doubts as to the extent that cable provided independent local news — the Commission’s recognized indicator of viewpoint diversity in local markets. Order ¶ 394 (“News and public affairs programming is the clearest example of programming that can provide viewpoint diversity .... [and] the appropriate geographic market for viewpoint diversity is local.”). While the responses to one question in the Commission’s survey suggested that cable is a significant source of local news, see MOWG Study No. 8 tbl. 8, the Commission did not find this credible because the responses to other survey questions suggested that respondents were counting broadcast signals that are transmitted as cable channels as sources of local news on cable. See id. tbl. 18 (from a list of popular national cable news channels and a choice for “local cable news channels,” almost half of the respondents chose “other” as their source for cable network news, indicating that cable news channels were probably confused with broadcast networks’ news). The survey’s indication that cable is an independent source of local news was also undercut by external evidence, including Neilsen ratings, that local cable channels are the least watched of any broadcast or cable stations in the market. Order ¶ 414. Furthermore, the Commission noted that local cable channels are not available everywhere, but only in select markets. Id. ¶ 414 n. 924; see also UCC Comments at 30-31 (reporting that only 10 to 15% of cable systems include channels that provide local and public affairs programming — i.e., public, educational, and governmental access channels — and that there are only 22 local news cable channels in the country, five of which serve the New York City area). Thus, the Commission justifiably excluded cable from its Diversity Index calculations.
Similarly, the responses to one of the Commission survey questions suggested that the Internet is a source of local news. See MOWG Study No. 831 tb1.1 (18.8% of respondents listed the Internet as a source of local news). But the survey did not identify which websites respondents used as sources of local news.32 There is a *406critical distinction between websites that are independent sources of local news and websites of local newspapers and broadcast stations that merely republish the information already being reported by the newspaper or broadcast station counterpart. The latter do not present an “independent” viewpoint and thus should not be considered as contributing diversity to local markets. Accordingly, the Commission should have discounted the respondents who primarily rely on these websites from its total number of respondents who indicated that they use the Internet to access local news.33
Furthermore, just as the Commission discounted responses indicating that cable was an independent source of local news because this finding conflicted with other record evidence, it should have discounted the Internet responses as well. The Commission does not cite, nor does the record contain, persuasive evidence that there is a significant presence of independent local news sites on the Internet. According to the record, most sources of local news on the Internet are the websites for newspapers and broadcast television stations. See, e.g., UCC Comments at 33 (62% of Internet users get local news from newspaper websites, 39% visit television station websites). And the examples the Commission does cite — the Drudge Report and Salon.com — have a national, not local, news focus.34 Order ¶ 427.
The Commission suggests that “the virtual universe of information sources” on the Internet qualifies it as a source of viewpoint diversity. Order ¶ 427. But to accept this rationale we would have to distort the Commission’s own premise that local news is an indicator of viewpoint diversity. E.g., id. ¶ 391 (Diversity Index measures “the relative importance of these media as a source of local news”).35 Search-engine sponsored pages such as Yahoo! Local and about.com,36 which were *407suggested by eommenters as sources of local news and information, may be useful for finding restaurant reviews and concert schedules, but this is not the type of “news and public affairs programming” that the Commission said was “the clearest example of programming that can provide viewpoint diversity.” Id. ¶ 394.
To accept its “universe of information” characterization of the Internet’s viewpoint diversity, we would also have to disregard the Commission’s professed intent to focus its consideration of viewpoint diversity on media outlets. Id. ¶¶ 20, 391. In terms of content, “the media” provides (to different degrees, depending on the outlet) accuracy and depth in local news in a way that an individual posting in a chat room on a particular issue of local concern does not. But more importantly, media outlets have an entirely different character from individual or organizations’ websites and thus contribute to viewpoint diversity in an entirely different way. They provide an ag-gregator function (bringing news/information to one place) as well as a distillation function (making a judgment as to what is interesting, important, entertaining, etc.).37 Individuals (such as political candidates) and entities (such as local governments or community organizations) may use the Internet to disseminate information and opinions about matters of local concern (such as the extension of a bike path on the Schuylkill River in Center City Philadelphia, see dissent Part IV.A.1), but these individuals and organizations are not, themselves, media outlets. We agree that the Internet “helps citizens discharge the obligations of citizenship in a democracy,” Order ¶ 393, when someone can go cityof-glenfalls.com to find the city council’s next agenda or use sfgov.org to learn how to get a marriage license in San Francisco. See Comments of the Hearst Corporation, MM Docket No. 01-235 at 11 & n. 36, apps. A & C (Dec. 3, 2001) (“Hearst Comments”) (listing these website addresses, among others, as examples of local government websites). But local governments are not, themselves, “media outlets” for viewpoint-diversity purposes. Like many entities, they just happen to use a particular media outlet — the Internet — to disseminate information.
Similarly, advertiser-driven websites such as hvnet.com and sfadvertiser.com, see Hearst Comments at apps. A & C, hardly contribute to viewpoint diversity. Like local governments, the sponsors of these websites are not “media outlets” just because they have “local information” that they want the public to have — if they were, their advertisements in telephone directories would also have to count toward viewpoint diversity. Compare Order ¶ 424 (explaining that the Commission would not count toward viewpoint diversity a local newspaper’s comics and classified ads because, while the subscriber is “undoubtedly getting a valuable service, it is not clear that the service has anything to do with news and current affairs”).
The Commission attempts to justify different treatment for cable and the Internet by suggesting that local cable news channels are only available in select markets, while the Internet is available everywhere. Not only is this distinction demonstrably false (as even the Commission acknowledged that almost 30% of Americans do not have Internet access, Order ¶ 365), it is irrelevant. That the Internet *408is more available than local cable news channels does not mean that it is providing independent local news. On remand the Commission must either exclude the Internet from the media selected for inclusion in the Diversity Index or provide a better explanation for why it is included in light of the exclusion of cable. See Sinclair, 284 F.3d at 163-65 (unexplained inconsistency was arbitrary and capricious).
3. The Commission did not justify its assumption of equal market shares.
Both the Citizen Petitioners and the Deregulatory Petitioners object to the Commission’s decision to assign all outlets within the same media type (that is, television stations, daily papers, or radio stations) an equal market share. The assumption of equal market shares is inconsistent with the Commission’s overall approach to its Diversity Index and also makes unrealistic assumptions about media outlets’ relative contributions to viewpoint diversity in local markets.38
The Commission’s decision to assign equal market shares to outlets within a media type does not jibe with the Commission’s decision to assign relative weights to the different media types themselves, about which it said “we have no reason to believe that all media are of equal importance.” Order ¶ 409; see also id. ¶ 445 (“Not all voices, however, speak with the same volume.”). It also negates the Commission’s proffered rationale for using the HHI formula in the first place — to allow it to measure the actual loss of diversity from consolidation by taking into account the actual “diversity importance” of the merging parties, something it could not do with a simple “voices” test. Id. ¶ 396. Finally, assigning equal market shares to outlets that provide no local news almost certainly presents an understated view of concentration in several markets, thus contravening the Commission’s goal of making “the most conservative assumption possible” about viewpoint diversity. Id. ¶ 400.
Additionally, there is no dispute that the assignment of equal market shares generates absurd results. For example, in New York City, the Dutchess Community College television station and the stations owned by ABC each receive an equal 4.3% market share. Or compare the Dutchess Community College station’s weighted share of 1.5% (4.3% times the 33.8% multiplier for television) to a mere 1.4% weighted, combined share assigned to the New York Times Company’s co-owned daily newspaper and radio station. Order app. C. A Diversity Index that requires us to accept that a community college television station makes a greater contribution to viewpoint diversity than a conglomerate that includes the third-largest newspaper in America39 also requires us to abandon both logic and reality.
The Commission’s attempt to justify its failure to consider actual market share of outlets within a media type is not persua*409sive. It suggests that actual-use data use data is not relevant because “current behavior is not necessarily an accurate predictor of future behavior.” Order ¶ 423. But this truism did not prevent the Commission from preferring actual-use data in assigning relative weight to the different media types. The Commission also suggests that, compared to consumer preferences for a general media type (which are generally stable), consumers’ preferences for particular media outlets are fluid because they depend on the media outlet’s chosen format or quantity or quality of content, which are easily changed. Id. ¶ 422. But while variance in the local news content of individual media outlets is conceivable — e.g., the home shopping television station could start carrying local news — -the Commission does not provide any evidence that media outlets acttially undergo any such radical content change, let alone that they regularly do so. Simply put, the Commission needs to under-gird its predictive judgment that stations can freely change the level of their news content with some evidence for that judgment to survive arbitrary and capricious review. Fox I, 280 F.3d at 1051.
Lastly, the Commission attempts to justify its refusal to employ actual-use data by arguing that collecting “information on viewing/listening/reading of local news and current affairs material” would make it “necessary first to determine which programming constituted news and current affairs,” which in turn “would present both legal/Constitutional and data collection problems.” Order ¶ 424. With respect to “legal/Constitutional” problems, the Commission is apparently concerned about categorizing programming as news or “non-news.” But the Commission obtained actual-use data in MOWG Study No. 8 without any “legal/Constitutional problems.” There it avoided having to make content-distinguishing judgments simply by asking respondents where they got their local news. And the Commission’s reference here to data collection problems is vague and unexplained; there is no suggestion that obtaining actual-use data for outlets within a media type would be prohibitively more onerous than obtaining the same data for the media types themselves, as it did in MOWG Study No. 8.
Because the Commission’s reasons for eschewing actual-use data in assigning market shares to outlets within a media type and assuming equal market shares are unrealistic and inconsistent with the Commission’s overall approach to the Diversity Index and its proffered rationale, we remand for the Commission’s additional consideration of this aspect of the Order.
4. The Commission did not rationally derive its Cross-Media Limits from the Diversity Index results.
After the Commission calculated the average Diversity Index scores for markets of various sizes, it set out to determine whether the increase in those scores resulting from different consolidation scenarios would have an acceptable or unacceptable effect. The Commission’s results, contained in appendix D of the Order, are replicated here for ease of reference. (We have used boldface type to indicate those increases in Diversity Index scores that the Commission found acceptable in crafting the Cross-Media Limits.)
*410[[Image here]]
As illustrated by the chart’s figures that are not in boldface type, the Commission prohibited all cross-media consolidation in the smallest markets where, under any of their sample consolidation scenarios,40 the increase in Diversity Index would be excessive. The Commission decided to permit all of the scenarios in the largest markets, finding the Diversity Index increases to be acceptable. In the mid-sized markets, the Commission allowed most of the consolidation scenarios but prohibited consolidations involving a newspaper and a television duopoly.
The Deregulatory Petitioners suggest that the Commission should have prohibited only cross-ownership mergers that would lead to an increase of more than 400 points, consistent with the practice of the Department of Justice to block mergers that would increase the local market’s HHI score by over 400 points. As illustrated by the chart above, this approach would have generated far more permissive Cross-Media Limits. The Citizen Petitioners argue, on the other hand, that the Commission should have prohibited cross-ownership mergers that lead to an increase of more than 100 points, the increase that triggers the Department of Justice’s “further review.” But the decision of where to draw the line between acceptable increases and unacceptable increases is almost always the Commission’s to make. Deference to the Commission’s judgment is highest when assessing the rationality of the agency’s line-drawing en*411deavors. See NCCB, 436 U.S. at 814-15, 98 S.Ct. 2096; AT&T Corp., 220 F.3d at 627. The Commission rationalized its decision to make conservative assumptions in order to “protect[its] core policy objective of viewpoint diversity.” Order ¶ 399; see also Sinclair, 284 F.3d at 153 (Commission determined that antitrust merger guidelines “might be too low as their purpose lay in defining the point at which antitrust scrutiny is required, and not in encouraging a wide array of voices and viewpoints”).
Although the Commission is entitled to deference in deciding where to draw the line between acceptable and unacceptable increases in markets’ Diversity Index scores, we do not affirm the seemingly inconsistent manner in which the line was drawn. As the chart above illustrates, the Cross-Media Limits allow some combinations where the increases in Diversity Index scores were generally higher than for other combinations that were not allowed. Consider the mid-sized markets (four to eight stations), where the Commission found that a combination of a newspaper, a television station, and half the radio stations allowed under the local radio rule would increase the average Diversity Index scores in those markets by 408 (four stations), 393 (five), 340(six), 247 (seven), and 314 (eight) points respectively. These permitted increases seem to belong on the other side of the Commission’s line. They are considerably higher than the Diversity Index score increases resulting from other combinations that the Commission permitted, such as the newspaper and television combination, 242 (four stations), 223 (five), 200(six), 121 (seven), and 152 (eight). They are even higher than those resulting from the combination of a newspaper and television duopoly — 376 (five stations), 357(six), 242 (seven), and 308 (eight)— which the Commission did not permit. The Commission’s failure to provide any explanation for this glaring inconsistency is without doubt arbitrary and capricious, and so provides further basis for remand of the Cross-Media Limits.41
5. The Commission should provide better notice on remand.
Our decision to remand the Cross-Media Limits also gives the Commission an opportunity to cure its questionable notice. Under the APA, an agency must publish notice of either the terms or substance of the proposed rule or a description of the subjects and issues involved. 5 U.S.C. § 553(b)(3). We have held that “the adequacy of the notice must be tested by determining whether it would fairly apprise interested persons of the ‘subjects and issues’ before the agency.” Am. Iron & Steel Inst. v. EPA, 568 F.2d 284, 293 (3d Cir.1977). The Citizen Petitioners argue that, in order to “fairly apprise” the public, the Commission is also obligated to provide notice of its underlying methodology, the reasoning from which it derived the proposed rule. The Commission’s notice *412only indicated that it was considering “creating a new metric” to “reformulate [its] mechanism for measuring diversity and competition in a market,” and that it was contemplating “designing] a test that accords different weights to different outlet types.” Notice, 17 F.C.C.R. 18,503, ¶¶ 113-15. The Citizen Petitioners argue that the Commission should have publieally noticed the Diversity Index once it determined that was the methodology on which it would rely to derive the Cross-Media Limits. See McLouth Steel Prods. Corp. v. Thomas, 838 F.2d 1317 (D.C.Cir.1988) (failure to describe a particular “model” for computing contamination levels was not adequate notice); United States v. Nova Scotia Food Prods. Corp., 568 F.2d 240, 251 (2d Cir.1977) (agency’s failure to provide notice of the data from which it derived regulation foreclosed “criticism of the methodology used or the meaning to be inferred from the data”).
The Commission argues that the Diversity Index was formulated as a response to comments, and it need not seek additional public comment on data formulated as a response to earlier comments.42 We are mindful that the APA’s notice obligations are not supposed to result in a notice-and-eomment “revolving door.” “Rulemaking proceedings would never end if an agency’s response to comments must always be made the subject of additional comments.” Cmty. Nutrition Inst. v. Block, 749 F.2d 50, 58 (D.C.Cir.1984). But courts have also suggested that an agency may withhold notice of its comment-derived data only in the absence of prejudice. Id. (“The response may, moreover, take the form of new scientific studies without entailing the procedural consequences appellants would impose, unless prejudice is shown.”); see also Solite Corp. v. EPA, 952 F.2d 473, 484 (D.C.Cir.1991) (allowing agency to use “ ‘supplementary’ data, unavailable during the notice and comment period, that ‘expand[s] on and confirm[s]’ information contained in the proposed rulemaking and addresses ‘alleged deficiencies’ in the preexisting data, so long as no prejudice is shown” (citing Cmty. Nutrition Inst., 749 F.2d at 57-58) (alteration in original)). As the Diversity Index’s numerous flaws make apparent, the Commission’s decision to withhold it from public scrutiny was not without prejudice. As the Commission reconsiders its Cross-Media Limits on remand, it is advisable that any new “metric” for measuring diversity and competition in a market be made subject to public notice and comment before it is incorporated into a final rule.
Y. Local Television Ownership Rule
Both the Citizen Petitioners and the De-regulatory Petitioners challenge the Commission’s modification to the local television ownership rule, which would allow triopolies in markets of 18 stations or more and duopolies in other markets, subject to a restriction on combinations of the four largest stations in any market. We uphold the top-four restriction but remand the numerical limits for the Commission to harmonize certain inconsistencies and better support its assumptions and rationale. We also remand for the Commission to reconsider or justify its decision to expand the rule’s waiver provision — applicable to sales of failed, failing, or unbuilt television stations — by eliminating the requirement that waiver applicants notice the station’s availability to out-of-market buyers.
*413A. Regulatory Background and the 2002 Biennial Review
As part of the 1996 Act, Congress directed the Commission to conduct a rule-making to determine whether to “retain, modify, or eliminate” its local television ownership rule, which at the time prohibited the common ownership of two television stations with overlapping Grade B signal contours.43 § 202(c)(2), 110 Stat. 111; Amendment of Sections 73.35, 73.21.0, and 73.636 of the Commission’s Rules Relating to Multiple Ownership of Standard, FM and Television Broadcast Stations, 45 F.C.C. 1476, ¶ 3 (1964) (establishing the restriction on the common ownership of televisions stations).
In response to Congress’s directive, the Commission promulgated a rule allowing an entity to own two television stations in the same DMA, provided that: (1) the Grade B field-strength contours of the stations do not overlap; and (2)(a) at least one of the stations is not ranked among the four highest-ranked stations in the DMA, and (b) at least eight “voices” — that is, independently owned, operational, commercial or noncommercial full-power broadcast television stations — would remain in the DMA after the proposed combination. 1999 Television Rule Review, 14 F.C.C.R. 12,903, ¶ 8; 47 C.F.R. § 73.3555(b). The D.C. Circuit Court reviewed this so-called “duopoly rule,” and in 2002 remanded it for the Commission to justify its decision to count only broadcast television stations as voices to the exclusion of other non-broadcast media. Sinclair, 284 F.3d at 162.
The Commission consolidated Sinclair's remand order with its 2002 biennial review under § 202(h). The Order announced the Commission’s decision to abandon the duopoly rule in favor of a rule that would permit common ownership of two commercial stations in markets that have 17 or fewer full-power commercial and noncommercial stations, and common ownership of three commercial stations in markets that have 18 or more stations. Both limits are subject to a restriction44 on the common ownership of stations ranked among the market’s largest four based on audience share. Order ¶ 186. The Commission also decided to expand its criteria for waiving the rule’s ownership restrictions for proposed combinations that involve failed, failing, or unbuilt stations by eliminating the requirement that waiver applicants provide notice of the sale to out-of-market buyers. Id. ¶ 225.
B. We uphold several threshold challenges to the Commission’s overall regulatory approach.
Before we consider the specific modifications to the local television rule, we address some threshold challenges to the Commission’s regulatory approach.
1. Limiting local television station ownership is not duplicative of antitrust regulation.
We reject the Deregulatory Petitioners’ contention that the Commission’s *414local television rule is duplicative of antitrust enforcement (by the Department of Justice and the Federal Trade Commission) and, thus, not in the public interest. The Commission ensures that license transfers serve public goals of diversity, competition, and localism, while the antitrust authorities have a different purpose: ensuring that merging companies do not raise prices above competitive levels. See, e.g., Clayton Act, § 7, 15 U.S.C. § 18 (restraining mergers that would lessen competition in a market); Dep’t of Justice and Federal Trade Comm’n, Horizontal Merger Guidelines § 0.1 (1997 rev. ed.) (“Merger Guidelines”) (seeking to protect consumers by ensuring mergers do not result in anticompetitive prices).
The Deregulatory Petitioners contend that the Commission’s new regulations are not derived from its professed reliance on audience preference, which the Commission suggests distinguishes its regulatory approach from that of the antitrust authorities. Order ¶ 65 (pointing out that it considers audience preferences plus advertising data as indicators of competition, while the antitrust authorities focus on prices). But the Commission’s local television ownership rule does reflect audience preferences in at least three ways. First, the Commission decided to focus its competition analysis on the delivered program market (as opposed to the video programming market or the video advertising market) because that is the market that “directly affects viewers.” Id. ¶ 141. Second, as discussed in Part V.C below, the Commission used audience preference data to support its conclusion that common ownership of local television stations can improve program quality. Id. ¶ 150. Third, as also discussed in Part V.C, the Commission justified the top-four restriction on evidence of an audience share “cushion” between the top-four stations and the fifth-ranked station in most markets. Id. ¶ 195. Thus, we reject the De-regulatory Petitioners’ suggestion that the local television ownership rule does not reflect the Commission’s concern for audience preferences.
Finally, we note that the Commission reviews all license transfers, 47 U.S.C. § 310(d), while the antitrust agencies typically review only large mergers. See 15 U.S.C. § 18a(a) (a merger must only be reported to the FTC and DOJ if the size of the transaction exceeds $200 million or if the assets of one party exceed $10 million and the assets of the other party exceed $100 million). Eighty-five percent of station mergers that have taken place since 2000 would not have been subject to antitrust review because the parties’ assets fell below these thresholds. Br. of Intervenor UCC at 37 (citing BIA Financial Network, Television Market Report (2d ed. 2003)). In this context, it hardly seems that the Commission’s local television station ownership rule is duplicative of other agencies’ antitrust enforcement.
2. Media other than broadcast television may contribute to viewpoint diversity in local markets.
Recognizing that allowing more television concentration in local markets could detract from viewpoint diversity, the Commission rationalized its decision to deregulate with its finding that media other than broadcast television contribute to viewpoint diversity. Order ¶ 171. This is a departure from the Commission’s rationale for the existing rule — the issue remanded by the Sinclair Court — that only television stations are relevant to its diversity analysis.
We agree with the Commission’s conclusion that broadcast media are not the only media outlets contributing to viewpoint diversity in local markets. Yet because we *415remand the Commission’s numerical limits, as explained in Part V.D below, we need not decide the degree to which non-broadcast media compensate for lost viewpoint diversity to justify the modified rule. Rather, we leave it for the Commission to demonstrate that there is ample substitut-ability from non-broadcast media to warrant the particular numerical limits that it chooses on remand.
We note that the record contains only weak evidence that cable can substitute for broadcast television as a source of viewpoint diversity. For example, the Commission found that among cable subscribers (a class that already omits one-third of American households) only 30% have access to local cable news channels. See Order ¶ 414 n. 924; see also supra Part IV.D.2 (describing why the Commission concluded that a majority of survey respondents’ purported reliance on cable as a source of local news was probably based on confusion with broadcast news); Reply Comments of the UCC, MM Docket No. 01-235 at 5 (Feb. 15, 2002), UCC Comments at 30-31. With respect to the Internet, while record evidence indicates a negative correlation between respondents’ reliance on broadcast television and the Internet as news sources (suggesting that people who use the Internet for local news do so at the expense of television), the Internet is also limited in its availability and as a source of local news. Compare MOWG Study No. 8 at 3, 34; with UCC Reply Comments at 10 and MOWG Study No. 8 tbl. 097. Therefore, it seems that the degree to which the Commission can rely on cable or the Internet to mitigate the threat that local station consolidations pose to viewpoint diversity is limited.
3. Consolidation can improve local programming.
The Commission supported its decision to relax the existing “eight voices” rule with findings that common ownership of television stations in local markets can result in “consumer welfare enhancing efficiencies” by eliminating redundant expenses and increasing opportunities for cross-promotion and related programming. Order ¶ 147. To promote localism, the Commission found that these efficiencies translated into improved local news and public interest programming.45 Id. ¶ 164. Evidence supporting this conclusion included findings that commonly owned television stations are more likely to carry local news than other stations and air a similar quality and quantity of local news as other stations. See Bruce M. Owen et al., Effect of Common Ownership or Operation on Television News Carriage, Quantity and Quality, in Comments of Fox Entertainment Group, Inc. et al., MB Docket 02-277 (Jan. 2, 2003). And a study of seven commonly owned broadcast television stations indicates that consolidation generally improved audience ratings — the studied stations increased their audience shares by an average of 3.2% over the share they enjoyed prior to entering into co-ownership with another station. Order ¶ 150 & n. 295 (citing Mark R. Fratrik, Television Local Marketing Agreement and Local Duopolies; Do They Generate New Competition and Diversity? (Jan. 2003), appendix to Comments of Coalition Broadcasters et al., MM Docket 02-277 (Jan. 2, 2003)). In light of this evidence, we reject the Citizen Petitioners’ conten*416tion that the Commission’s finding of local-ism benefits from consolidation was unsupported.
4. The Commission adequately noticed its decision to allow triopo-lies.
We also reject the Citizen Petitioners’ contention that the Commission failed to provide adequate notice that it was considering a rule that would allow triopolies. The APA requires agencies to publish a notice of proposed rulemaking that contains “either the terms or substance of the proposed rule or a description of the subjects and issues involved.” 5 U.S.C. § 553(b). The Notice advised parties that any reevaluation of the television ownership rule would take account of the remand ordered by Sinclair, which left it to the Commission to reexamine its “voices” test as well as “the numerical limit, given that there is a relationship between the definition of voices and the choice of a numerical limit.” Notice, 17 F.C.C.R. 18,503, ¶ 76 (citing Sinclair, 284 F.3d at 162). Specifically, the Commission asked for comment on “different economic incentives” relating to diverse viewpoints in newscasting that might exist “among stand-alone stations, duopolies, or triopolies.” Id. ¶ 80. This leaves little doubt that the Notice provided a sufficient “description of the subjects and issues involved” in the Commission’s decision to allow triopolies.
C. We uphold the Commission’s decision to retain the top-four restriction.
Though the Commission recognized that the combination of television stations within the same market could yield efficiencies that benefit consumers, the Commission also recognized that station combinations only have an overall public “welfare-enhancing” effect when the consolidation does not create a “new largest” entity. Order ¶ 194 (citing, inter alia, R. Preston McAfee & Michael Williams, Horizontal Mergers and Antitrust Policy, 40 J. Indus. Econ. 181-87 (June 1992)). Thus, the Commission determined that it had to limit allowable station combinations to those that would not create excessive market power in a “new largest” entity. Finding that a significant “cushion” of audience share percentage points generally separates top-four stations from the fifth-ranked stations, the Commission decided that a top-four restriction would ensure that station consolidations did not lead to excessive market power. Id. ¶ 195-96. The Commission also recognized that efficiencies are less prevalent when financially strong stations merge with each other. Id. ¶ 197. For example, top-four stations are already more likely to be originating local news46 and to have made the transition to digital television. Id. ¶¶ 198-99.
The Deregulatory Petitioners assert that the top-four restriction prevents small-market stations from yielding any of the benefits of consolidation. They argue that by effectively precluding any consolidation in markets with fewer than five stations, the Commission deprives the benefit of consolidation from stations — those in small markets — who need it most. But the Commission’s local television rule is protective of small-market stations. The numerical limits already allowed, in effect, *417extra47 concentration to ensure that small-market stations would realize the efficiency benefits of consolidation. As for the smallest markets with fewer than five stations — where the top-four restriction operates to preclude any consolidation — it was not unreasonable for the Commission to conclude, as it did, that the detriment of concentrated market power — e.g., reduced incentive to improve programming of mass appeal — outweighed the efficiency benefits. Id. ¶¶ 197, 200. And, while we recognize that the Commission “cannot save an irrational rule by tacking on a waiver procedure,” ALLTEL Corp. v. FCC, 888 F.2d 551, 561 (D.C.Cir.1988), we note that the modified rule allows the Commission to waive the top-four restriction in small markets where those consolidations would be beneficial overall.48
The Deregulatory Petitioners also object to the top-four restriction because, they argue, it unjustifiably treats all top-four ranked stations the same. The essence of their objection is the Commission’s finding of a “general separation” between the top-four stations and other stations. They point out that in many markets, especially small-and medium-sized ones,49 the third and fourth-ranked stations could combine without exceeding the audience share of the first or second-ranked stations. Thus, they contend, there is a more substantial “separation” between the third and fourth-ranked stations than between the fourth and fifth-ranked stations.50
But we must uphold an agency’s line-drawing decision when it is supported by the evidence in the record. Sinclair, 284 F.3d at 162; AT&T Corp., 220 F.3d at 627. Here there is ample evidence in the record *418to support the Commission’s restriction on combinations among the top-four stations as opposed to the top-three or some other number. The Commission found a “cushion” of audience share percentage points between the fourth and fifth-ranked stations in most markets. Order ¶ 195. Networks’ national audience statistics, which are generally reflected in local market rankings of affiliated stations, also show a substantial 60% drop in audience share between the fourth and fifth-ranked networks. In the ten largest markets, the top-four stations combined control at least 69% (and an average of 83%) of the local commercial share in their respective markets, and in all of the ten largest markets a combination between the third and fourth-ranked stations would produce a new largest station. Local Television Ownership and Market Concentration Study, in UCC Reply Comments at attachment 5.51 Furthermore, the Commission found that permitting mergers among top-four ranked stations generally leads to large HHI increases.
Thus we conclude that the Commission’s decision to retain the top-four restriction is supported by record evidence. Accordingly, we extend deference. See Sinclair, 284 F.3d at 162.
D. We remand the specific numerical limits for the Commission’s further consideration.
The Commission decided to construct its numerical limits to ensure that most markets would have six equal-sized competitors because the HHI score of a six-member market — 1667 (6 x (100 -e 6)2 ) — is below the Department of Justice and Federal Trade Commission’s 1800 threshold for highly concentrated markets for antitrust purposes. Order ¶¶ 192-93 (citing Merger Guidelines § 1.51(c)). Thus the Commission decided to allow triopolies in markets of 18 stations or more (18 -h- 3 = 6 equal-sized competitors) and duopolies in markets of 17 or fewer (both limits subject to the restriction on a combination of top-four stations). Id. ¶ 193.
The Commission’s assumption of equal market shares received flak from both ends of the objecting spectrum. The Citizen Petitioners point out that television stations’ market shares vary widely and argue that it is arbitrary for the Commission to base its numerical limits on a rudimentary station “head count” of outlets. The Commission’s rationale for its triopoly rule requires that we accept a combination of the first, fifth, and sixth-ranked stations as the competitive equal of a combination of the 16th, 17th, and 18th-ranked stations, just because each combination consists of the same number of stations. While the Citizen Petitioners demonstrate the Commission’s flawed rationale with examples of what the modified rule allows, the Deregu-latory Petitioners demonstrate the same flaw by pointing out what the modified rule forbids. There is no logical reason, they argue, why it should be impermissible to have five duopolies and one triopoly (a total of six competitors)52 in a market with *41913 stations when it is possible that the triopoly could have a lower combined market share than any or all of the duopolies.
The Commission defends its equal market shares approach with the suggestion that market share, which varies with each season’s new programs, is too “fluid” to be the basis for its regulations. Id. ¶ 193. But elsewhere in the local television ownership rule the Commission found that the market share was stable enough to rely on for support of its top-four restriction.53 Id. ¶ 195. And not only is the Commission’s “market share is too fluid” rationale inconsistent with other aspects of the rule, it is unsupported. The Order cites no evidence to support its assumption that market share fluctuates more in television-broadcasting than in other industries. Nor does it refute the Citizen Petitioners’ suggestion that this is unlikely to be the case because, unlike most other industries, television station owners face a barrier to market entry (requirement of a license) and the number of market participants (television station owners) is in decline.
The Citizen Petitioners also object to the' numerical limits because they allow markets that are already highly concentrated to become even more concentrated. For example, Philadelphia is a market "with more than 18 stations, so the modified rule would allow the duopolist Viacom to acquire a third station and potentially increase its audience share from 25% to 34%. Philadelphia’s HHI score is already 2037, well above the Commission’s 1800 target, and a Viacom triopoly would raise the score to 2487. We acknowledge that the Commission never intended the numerical limits to represent a “mechanical application of the DOJ/FTC Merger Guidelines.” Id. 11197. . But it expressly chose its six equal-sized competitor benchmark to ensure that markets would not exceed the Merger Guidelines’ 1800 threshold for highly competitive markets.54 Id. ¶ 192. After justifying 1800 as the target, the Commission relaxed the local television ownership rule to allow more concentration in markets that already exceed the target-which is not just some markets, but most. UCC Reply Comments attachment 5 (28 out of 33 markets studied had HHIs that exceeded 1800, based on 2001 data).55
*420The deference with which we review the Commission’s line-drawing decisions extends only so far as the line-drawing is consistent with the evidence or is not “patently unreasonable.” Sinclair, 284 F.3d at 162. The Commission’s numerical limits are neither. No evidence supports the Commission’s equal market share assumption, and no reasonable explanation underlies its decision to disregard actual market share. The modified rule is similarly unreasonable in allowing levels of concentration to exceed further its own benchmark for competition (1800)-a glaring inconsistency between rationale and result. We remand the numerical limits for the Commission to support and harmonize its rationale.
E. We remand the Commission’s repeal of the Failed Station Solicitation Rule.
The Citizen Petitioners also challenge the Commission’s repeal of the Failed Station Solicitation Rule (“FSSR”), 47 C.F.R. 73.3555 n.7, which required a waiver applicant to provide notice of the sale to potential out-of-market buyers before it could sell the failed, failing, or unbuilt television station to an in-market buyer.56 Order ¶ 225.
The Commission promulgated the FSSR in its review of the local television rule that Congress had required under § 202(c) of the 1996 Act. To alleviate concerns that its decision to allow duopolies would undermine television station ownership by minorities, the Commission created the FSSR to ensure that qualified minority broadcasters had a fair chance to learn that certain financially troubled — and consequently more affordable — stations were for sale. 1999 Television Rule Review, 14 F.C.C.R. 12,903, ¶¶ 13-14, 74. In the current Order, however, the Commission does not explain that preserving minority ownership was the purpose of the FSSR, nor does it argue that the FSSR was harmful or ineffective toward this purpose. The only support for its decision to eliminate the FSSR is its prediction that “the efficiencies associated with operation of two same-market stations, absent unusual circumstances, will always result in the buyer being the owner of another station in that market.”57 Order ¶ 225.
By failing to mention anything about the effect this change would have on potential minority station owners, the Commission *421has not provided “a reasoned analysis indicating that prior policies and standards are being deliberately changed, not casually ignored.” Greater Boston TV Corp. v. FCC, 444 F.2d 841, 852 (D.C.Cir.1970). Furthermore, while the Commission had promised in 1999 to “expand opportunities for minorities and women to enter the broadcast industry,” 1999 Television Rule Review, 14 F.C.C.R. 12,903, ¶ 14, the FSSR remained its only policy specifically aimed at fostering minority television station ownership. In repealing the FSSR without any discussion of the effect of its decision on minority television station ownership (and without ever acknowledging the decline in minority station ownership notwithstanding the FSSR), the Commission “entirely failed to consider an important aspect of the problem,” and this amounts to arbitrary and capricious rule-making.58 State Farm, 463 U.S. at 43, 103 S.Ct. 2856; see also Copps Dissenting, 18 F.C.C.R. at 13,970-71 (chastising the Commission for “fail[ing] to conduct rigorous analysis of today’s rules on minorities and women”); Adelstein Dissent, 18 F.C.C.R. at 13,997 (same). For correction of this omission, we remand.59
VI. Local Radio Ownership Rule
Petitioners challenge the Commission’s decision to modify its local radio ownership rule, which limits the number of commercial radio stations that a party may own in local markets of different sizes, 47 C.F.R. § 73.3555(a), by, among other changes, adopting a new method for determining the size of local markets. They also argue that the Commission failed to justify its decision to retain the rule’s specific numerical limits. We affirm the Commission’s decision to modify the rule (including modifying its method for determining local market size), but we agree that its decision to retain the numerical limits was arbitrary and capricious, and hence remand for the Commission’s further consideration.
A. Regulatory Background and the 2002 Biennial Review
In 1992 the Commission abandoned its one-to-a-market limit on radio station ownership and implemented tiered numerical limits that allowed for common ownership of as many as six (three AM and three FM) stations — but no more than a combined 25% of the market’s total audience share — in the largest markets of 40 or more commercial stations.60 In 1996 Con*422gress directed the Commission to relax the tiered limits even more, to allow for as many as eight radio stations (but no more than five in the same AM or FM service) to be commonly owned in the largest markets of 45 or more commercial stations.61 Additionally, Congress required that the Commission review these limits biennially and repeal or modify them as necessary in the public interest. Id. § 202(h), 110 Stat. at 111-12.
In its 1998 and 2000 biennial reviews, the Commission decided to retain the same numerical limits enacted by Congress after finding that they continued to be necessary in the public interest. 1998 Biennial Regulatory Review, 15 F.C.C.R. 11,058, ¶ 59, 2000 WL 791562 (2000); 2000 Biennial Regulatory Review, 16 F.C.C.R. 1207, ¶ 32, 2001 WL 38123 (2001). But separately the Commission initiated a proceeding to consider modifying the way in which it “determine[s] the dimensions of radio markets and counts the number of stations in them.” Definition of Radio Markets, 15 F.C.C.R. 25,077, ¶ 1, 15 F.C.C.R. 25077 (2001). That rulemaking was later consolidated with the 2002 biennial review. Notice, 17 F.C.C.R. 18,503, ¶ 7.
In its 2002 biennial review, the Commission again decided that the numerical limits set by Congress in 1996 remained necessary in the public interest. Order ¶ 239. But the Commission changed its method for determining the size of a radio station market. Since the Commission first promulgated tiered numerical limits in 1992, it had defined the markets as the overlapping area of. radio stations’ coverage. 47 C.F.R. § 73.3555(a)(3)(h); 57 Fed.Reg. 18,089 (Apr. 29, 1992). For each proposed combination of stations, the local market was defined by “the number of ... stations whose principal community contours overlap, in whole or in part, with the principal community contours of the stations” that were proposed to be commonly owned. 47 C.F.R. § 73.3555(a)(3)(h).62
Following the 2002 biennial review, the Commission decided to replace this so-called “contour-overlap methodology” with the geography-based market definition used by Arbitron, a private entity that measures local radio station audiences for its customer stations. Order ¶¶ 275-76. Within these so-called “Arbitron Metro markets,” the number of total radio stations would include both commercial and noncommercial stations that are licensed to a community within the market, as well as stations located outside the market that attract a minimum level of listenership within, see id. ¶¶ 279-81, 295. Under the 1996 Act, only commercial stations had counted in determining the size of the market.
Though the Commission decided to grandfather any existing radio station combinations rendered noncompliant by the switch to Arbitron Metro markets, it restricted transfer of those combinations, subject to a limited exception for sales to eligible small businesses. Id. ¶ 489.
In an additional modification to the local radio ownership rule, stations under Joint Sales Agreements — as noted earlier, when *423a licensee authorizes a broker station to sell advertising time on the licensee’s station in return for a fee — would be “attributed” to the brokering entity for the purpose of determining the brokering entity’s compliance with the applicable limit. Id. ¶ 317.
B. We uphold the Commission’s new definition of local markets.
Both the Deregulatory Petitioners and the Citizen Petitioners object to the Commission’s changes to the way in which it defines a local radio market. Predictably, the Deregulatory Petitioners object to the modification that decreases the' size of local markets (using Arbitron Metro markets instead of the contour-overlap methodology) — which potentially lowers the applicable numerical limit in a given market — while the Citizen Petitioners object to the modification that increases the size of local markets (inclusion of noncommercial stations) — which potentially raises the applicable numerical limit in a given market. But we conclude that the Commission has justified these changes under § 202(h) and they are not arbitrary and capricious under the APA.
1. The Commission justified using Arbitron Metro markets.
We reject out of hand the Deregulatory Petitioners’ argument that, by changing the market definition methodology, the Commission effectively increased the numerical limits in violation of § 202(h)’s presumption in favor of deregulation. As discussed in Part II above, § 202(h) is not a one-way ratchet. The Commission is free to regulate or deregulate as long as its regulations are in the public interest and are supported by a reasoned analysis. The change to Arbitron Metro markets is both.
Under the existing rule, to determine whether an entity may acquire a radio station under the local radio rule, the Commission first must know how many radio stations are in that station’s local market (called the “denominator” figure). The size of the market determines which numerical limit applies. Second, the Commission must determine how many radio stations in that market would be owned by the same entity if the entity acquired the stations it proposes (called the “numerator” figure). If this figure is within the numerical limit, the transaction may proceed.
Under the contour-overlap methodology, the Commission calculates the numerator by counting the acquiring entity’s radio stations that all have overlapping signal contours. A station whose signal contour overlaps with some but not all of the other stations’ contours will not be counted in the numerator. The Commission calculates the denominator by counting all of the stations whose contours intersect with at least one (not all) of the contours of another station in the numerator.
The Commission found that the contour-overlap methodology results in inconsistency between the numerator and denominator figures for a given transaction. Some radio stations owned by the acquiring entity may be counted in the denominator (because their contours intersect with at least one of the numerator-station’s contours) but not the numerator (because their contours do not intersect with all of the other numerator-station’s contours). An acquiring entity might actually own stations that do not count toward determining how many more it may own within the numerical limit, but do count toward determining the size of the market, where a bigger number correlates to a higher applicable limit. Order ¶ 253. The numerator/denominator inconsistency gives *424rise not only to an artificially low numerator/denominator ratio, which favors combinations but also the potential that combination owners could strategically acquire stations — those whose signals overlap with some, but not all of their already-owned stations’ signals — to end-run the numerical limits altogether. Id. ¶ 254. Furthermore, a “perverse incentive” problem arises, as the contour-overlap methodology actually encourages consolidation of powerful radio stations because the stations with larger signals are more likely to create larger markets, which makes it more likely that their owner would be able to acquire even more radio stations in the market. Id. ¶ 257.
The Commission determined that it could not fix these problems with the contour-overlap methodology “merely by excluding commonly owned stations from the denominator or including those stations in the numerator.” Id. ¶ 255. Excluding those stations from the denominator would mean the Commission was determining the size of the market based on who owns the stations, “a distinction that would be both unprincipled and unprecedented in the history of competition analysis.” Id. On the other hand, including all commonly owned stations that are represented in the denominator could overly inflate ownership levels by including “outlying stations far from the area of concentration.” Id.
The Commission also found that its transaction-specific market definitions frustrated its ability to benchmark and compute the level of competition in a given area. Thus it determined that “a local radio market that is objectively determined, ie., that is independent of the radio stations involved in a particular acquisition, presents the most rational basis for the defining radio markets.” Id. ¶ 273. For these reasons, the Commission decided to jettison the contour-based methodology and instead employ Arbitron’s existing geographical definitions of local markets, finding they are the “established industry standard” that “represent a reasonable geographic market delineation within which radio stations compete.” Id. ¶ 276.63
As the Deregulatory Petitioners point out, the Commission’s reliance on Arbitron Metro markets is not without flaw. One obvious problem is that this method of measure does not cover the entire country. In fact, about 70% of counties in the United States are not within one of the 287 Arbitron Metro markets. But people tend to be clustered in specific population centers, as 78% of the United States’ population over the age of 12 lives in Arbitron Metro markets. Nor has the Commission ignored the minority of the public who does not live in Arbitron Metro markets. It initiated a new rulemaking proceeding to develop market definitions for these areas, and in the interim decided to use a modified contour-overlap methodology that “minimize[s] the more problematic aspects of that system.” Id. ¶ 285.
The Deregulatory Petitioners argue that the Commission failed to demonstrate that the contour-overlap methodology was actually harming competition. One petitioner called the Commission’s explanation of the numerator/denominator inconsistency an “exercise in graduate school theoretical economics.” Tr. at 394. But the Commission did find evidence of potential competitive harm in a working group study suggesting that station consolidation under the contour-overlap regime has resulted in an increase in advertising prices. Order ¶ 261 n. 548. And, as described above, the Commission observed that the eontour-ov-*425erlap methodology created a “perverse incentive” problem, as the most powerful stations whose contours intersect more stations (and thus have a larger denominator) are more able than smaller stations to acquire more stations. Id. ¶ 257. The Commission also found that the contour-overlap methodology impairs its ability accurately to measure and compare competition in consistently defined markets. Id. ¶ 259. The Commission’s findings provide ample justification for the conclusion that the contour-overlap methodology’s resulting inconsistencies are more than just abstract theory.
Next we reject the Deregulatory Petitioners’ argument that the Commission improperly departed from its past precedent, namely, its 1992 and 1994 decisions to retain the contour-overlap methodology. See Revision of Radio Rules and Policies, 7 F.C.C.R. 6387, ¶¶ 37-40, 1992 WL 690638 (1992); Revision of Radio Rides and Policies, 9 F.C.C.R. 7183, ¶35, 1994 WL 617591 (1994). In State Farm, the Supreme Court said that when an agency reverses its “former views” it is “obligated to supply a reasoned analysis for the change beyond that which may be required when an agency does not act in the first instance.” 463 U.S. at 41-42, 103 S.Ct. 2856; see also Rust v. Sullivan, 500 U.S. 173, 186, 111 S.Ct. 1759, 114 L.Ed.2d 233 (1991). Here the Commission supplied a reasoned analysis for the change — that “[i]t was only after the ownership limits were substantially raised in the 1996 Act that the scope of the market distorting effects of that system became manifest.” Order ¶ 262.
The Deregulatory Petitioners argue that because radio is a signal-based enterprise, any market definition that ignores the actual reach of station signals and the scope of their listenership does not accurately measure competition. But we believe that it was reasonable for the Commission to conclude that fixed, geography-based market definitions more readily enable accurate measurement and comparison of competition than a transaction-specific, contour-based definition. Id. ¶ 259.
Finally, the Deregulatory Petitioners point out that Arbitren makes money from its station customers, and it has financial and other obligations to those stations and its shareholders. Therefore, they argue, Arbitron’s economic incentives render its market definitions subject to manipulation. But the Commission recognized these potential problems and adequately established specific safeguards to deter potential manipulation, including a two-year buffer period before any party can receive the benefit of either a change in Arbitren Metro market boundaries or the addition of more radio stations to the market. Id. ¶ 278. The Deregulatory Petitioners’ suggested conjectural flaws do not persuade us that the Commission’s adoption of Arbi-tren Metro markets was arbitrary and capricious. We also reject their contention that the Commission’s use of Arbitren markets delegates governmental power to a private entity. Arbitren will only provide a mechanism for measuring concentration. Because the Commission remains the sole arbiter of whether a proposed radio station combination serves the public interest, no improper delegation will occur.
For these reasons, we conclude that the Commission’s decision to replace contour-overlap methodology with Arbitren Metro markets was “in the public interest” within the meaning of § 202(h) and that it was a rational exercise of rulemaking authority.
2. The Commission justified including noncommercial stations.
The Citizen Petitioners object to the Commission’s decision to count both noncommercial and commercial stations in *426determining the size of an Arbitron Metro market.64 They argue that the inclusion of noncommercial stations increases the number of stations in a given market and thereby allows more consolidation under the numerical limits. It is impermissibly inconsistent, they argue, to increase effectively the numerical limits in this way while at the same time defending the existing numerical limits from arguments that they should be relaxed. But we reject this argument’s factual premise that counting noncommercial stations will increase significantly the number of stations in a given market. Even counting noncommercial stations in the definition of markets, the change to Arbitron Metro markets operates as a net decrease in most markets’ size.65 This undercuts the Citizen Petitioners’ argument that it was inconsistent for the Commission to increase the sizes of local markets while justifying the existing numerical limits. It was not arbitrary and capricious, therefore, for the Commission to take account of noncommercial stations in determining the ownership limits.66
C. We uphold the Commission’s transfer restriction.
In applying its local ownership rules, the Commission decided to grandfather existing radio, television, and radio/television combinations, and did “not require entities to divest their current interests in stations in order to come into compliance with the new ownership rules.” Order ¶ 484.67 Under past Commission rules, radio stations created in compliance with the rules were freely transferable under the Commission’s policy that maintaining the status quo did not create new harm to public interest goals. See Revision of Radio Rules and Policies, 7 F.C.C.R. 6387, ¶ 48. Switching course, in this Order the Commission refused to permit the transfer or sale of grandfathered combinations that violate its local ownership limits except to certain “eligible entities” that qualify as *427small businesses.68 Order ¶ 488. We reject the Deregulatory Petitioners’ arguments that this modification to the local radio ownership rule contravenes § 202(h), constitutes arbitrary and capricious rule-making, and violates the United States Constitution.
1. Transfer restriction is “in the public interest.”
The Commission successfully demonstrates that the transfer restriction is “in the public interest” as required under § 202(h).69 The Commission notes that forcing divestitures at the time of transfer “could afford new entrants the opportunity to enter the media marketplace” and “could give smaller station owners already in the market the opportunity to acquire more stations,” which would promote the public interest in having competitive and diverse radio markets. Id. ¶ 487. Additionally, the Commission found a strong public interest in curtailing the perpetuation of “combinations [that] were created pursuant to a market definition that we conclude fails to adequately reflect competitive conditions.” Id. Unlike its decision not to require noncompliant combinations to divest, the Commission decided to restrict transfers because it found no “countervailing considerations” (such as the expectancy interests of existing station owners) to outweigh the public interest. Id. This rationale satisfies the public interest requirement under § 202(h).
2. Transfer restriction is reasoned decisionmaking.
According to the Deregulatory Petitioners, the transfer restriction is arbitrary and capricious because the Commission “failed to consider an important aspect of the problem” — as required under State Farm, see 463 U.S. at 43, 103 S.Ct. 2856—by ignoring the fact that combination-owners’ investments are diminished by the rule. Requiring station combinations to spin off noncompliant stations depresses the price of the station combinations as a whole, which deprives combination-owners (as sellers) of their expected investments. But an agency is not restricted in its rule-making by the expectations of the regulated. Borough of Columbia v. Surface Transp. Bd., 342 F.3d 222, 236 (3d Cir.2003) (recognizing that parties who enter the market knowing that the' industry is heavily regulated “cannot be said to have been surprised that the [agency’s] decisions might thwart its goals”); see also Sinclair, 284 F.3d at 166 (upholding a modification to the local television ownership rule “notwithstanding that it may upset some expectations”).
The Deregulatory Petitioners also suggest that the Order arbitrarily and capriciously exempted sales to “eligible entities” from the transfer restriction because the Commission did not purport to analyze whether small businesses ($6 million or less in annual revenue) would actually be able to obtain financing necessary to acquire a noncompliant radio station combination. But the rationality of the transfer restriction does not depend on the “eligible entities” exception. It is enough for the Commission to decide, as it did, that when it comes to small business the relative *428diversity advantages to a market in which an eligible-entity transfer occurred would outweigh the disadvantages to the market’s competition interests from the perpetuation of a noncompliant combination. The number of times the exception is invoked is irrelevant to the reasonableness of the transfer restriction as a whole.70
3. Transfer restriction is constitutional.
We reject also the Deregulatory Petitioners’ suggestion that the Commission’s restriction on the transferability of pre-existing noncompliant consolidations “raises serious constitutional questions”71 under the Due Process and Takings Clauses of the Constitution’s Fifth Amendment.72
In order to succeed in a due process or takings case under the Fifth Amendment, the plaintiff must first show that a legally cognizable property interest is affected by the Government’s action in question. See Cleveland Bd. of Educ. v. Loudermill, 470 U.S. 532, 538, 105 S.Ct. 1487, 84 L.Ed.2d 494 (1985) (requiring protected property interest for due process violation); Webb’s Fabulous Pharmacies, Inc. v. Beckwith, 449 U.S. 155, 160-61, 101 S.Ct. 446, 66 L.Ed.2d 358 (1980) (requiring protected property interest for taking). But broadcast licenses, which are the subject' of the Commission’s restriction on transferability, are not protected property interests under the Fifth Amendment. See FCC v. Sanders Bros. Radio Station, 309 U.S. 470, 475, 60 S.Ct. 693, 84 L.Ed. 869 (1940) (“The policy of the [1936 Communications] Act is clear that no person is to have anything in the nature of a property right as a result of the granting of a license.”); see also CBS, Inc. v. FCC, 453 U.S. 367, 395, 101 S.Ct. 2813, 69 L.Ed.2d 706 (1981) (noting that the broadcast spectrum is part of the public domain and that a broadcaster merely has “use of a limited and valuable part of the public domain; when he accepts that franchise it is burdened by enforceable public obligations” (quoting Office of Communication of the United Church of Christ v. FCC, 359 F.2d 994, 1003 (D.C.Cir.1966))); 47 U.S.C. § 301 (broadcast licenses provide for the “use ... but not the ownership” of channels of communication). Broadcast licensees accept their licenses subject to the Commission’s power to regulate their transfer pursuant to “the public interest, convenience, and necessity.” 47 U.S.C. § 310(d). Because a legally cognizable property interest is not implicated by the Commission’s decision to restrict the transfer of preexisting, noncompliant station combina*429tions, a Fifth Amendment violation will not occur.
D. We affirm the attribution of Joint Sales Agreements.
Under a Joint Sales Agreement (“JSA”), a radio station authorizes a broker to sell advertising time on the station in exchange for a fee. In the Order, the Commission decided to count stations brokered under a JSA toward the brokering station’s permissible ownership totals (that is, to “attribute” JSAs), as long as (1) the brokering entity owns or has an attributable interest in one or more stations in the local market, and (2) the joint advertising sales amount to more than 15% of the brokered station’s advertising time per week. Order ¶ 317. The Commission justified this departure from its prior policy of nonattribution with its finding that in-’ market JSAs above the 15% threshold convey to the brokering entity a degree of “influence or control” sufficient to warrant limitation under the ownership rules — ie., a “realistic potential to affect [the brokered station’s] programming and other core operating decisions.”73 Id. ¶ 318. Furthermore, the brokering entity has the ability and the incentive to exercise market power because it controls the advertising revenue of the brokered stations and assumes the financial risks and rewards of advertising. Id. ¶ 320. The Commission gave parties to JSAs two years to end their agreements or otherwise come into compliance with the local ownership rules. Id. ¶ 325. We reject the Deregulatory Petitioners’ arguments that this modification to the local radio ownership rule contravenes § 202(h), constitutes arbitrary and capricious rulemaking, and violates the Constitution.
1. Attribution of JSAs is “necessary in the public interest.”
In the Order the Commission explained that, due to their potential to convey influence to brokering entities, JSAs “may undermine [the Commission’s] continuing interest in broadcast competition sufficiently to warrant limitation under the multiple ownership rule.” Id. ¶ 318. The Commission’s decision to modify its attribution policy to “reflect accurately the competitive conditions of today’s local radio markets,” id. ¶ 321, and thus prevent its local radio rule from being undermined, is a modification “in the public interest” under § 202(h).74
2. Attribution of JSAs is reasoned decisionmaking.
An agency that departs from its “former views” is “obligated to supply a reasoned analysis for the change beyond that which may be required in the first instance” in order to survive judicial scrutiny for compliance with the APA. State Farm, 463 U.S. at 41-42, 103 S.Ct. 2856. Here the Commission acknowledged that its decision to attribute JSAs was a departure from its prior policy of nonattribution, promulgated in 1999. Order ¶ 320 n. 698. The Commission explained that, “upon reexamination of the issue,” it found that JSAs convey the potential for brokering entities to influence the brokered stations; thus a policy of attributing JSAs is necessary to “reflect accurately competitive conditions of today’s local radio markets.” Id. ¶¶ 320, 321.
*430The Deregulatory Petitioners argue that this proffered rationale is insufficient because the Commission does not explain what, if anything, transpired since 1999 that accounts for its change of position. But because we interpret the Commission’s modification as a correction of its prior policy’s “inaccuracy,” the Commission’s failure to cite a particular intervening change is not fatal to its reasoned analysis. In this context we accept that the Commission’s determination' — upon “reexamination of the issue” that the JSAs convey (and always have conveyed) a potential for influence — sufficiently rationalizes its decision to jettison its prior nonat-tribution policy and replace it with one that more accurately reflects the conditions of local markets.
3. Attribution of JSAs is constitutional.
We also disagree with the Deregulatory Petitioners’ suggestion that- the attribution of JSAs “raises serious constitutional concerns”75 under the Fifth Amendment’s Takings Clause. Contracts such as JSAs are protected property interests under the Fifth Amendment, see United States Trust Co. of N.Y. v. New Jersey, 431 U.S. 1, 19 n. 16, 97 S.Ct. 1505, 52 L.Ed.2d 92 (1977), but here the Commission has not invalidated or interfered with any contracts. In deciding to attribute JSAs, it is has simply decided that stations subject to JSAs should, in certain circumstances, count toward the regulatory limit in determining how many stations the brokering entity may own in a market. Moreover, station owners have no vested right in the continuation of any particular regulatory scheme. Folden v. United States, 56 Fed. Cl. 43, 61 (2003) (parties “in a highly regulated field such as FCC licensing can have no distinct investment-backed expectations that include a reliance upon a legislative and regulatory status quo”); cf. Connolly v. Pension Benefit Guar. Corp., 475 U.S. 211, 222-27, 106 S.Ct. 1018, 89 L.Ed.2d 166 (1986) (no regulatory taking occurred because government had not appropriated property for its own use, did not impose a severe ■economic impact, and did not interfere with reasonable expectations). Thus we reject the Deregulatory Petitioners’ suggestion that the attribution of JSAs will result in a regulatory taking.
E. We remand the numerical limits to the Commission for further justification.
In the Order, the Commission announced its decision to retain the existing numerical limits on radio ownership that Congress established in § 202(b) of the 1996 Act. The existing limits preclude the common ownership of radio stations with overlapping principal community contours, except that: (1) in a radio market with 45 or more commercial radio stations, a party may own up to eight commercial radio stations, not more than five of which are in the same service (that is, AM or FM); (2) in a radio market with between 30 and 44 commercial radio stations, a party may own up to seven commercial radio stations, not more than four of which are in the same service; (3) in a radio market with between 15 and 29 commercial radio stations, a party may own up to six commercial stations, not more than four of which are in the same service; and (4) in a market with 14 or fewer commercial radio stations, a party may own up to five commercial radio stations, not more than three *431of which are in the same service, except that a party may not own more than 50 percent of the stations in that market. 1996 Act, § 202(b)(1), 110 Stat. at 110; 47 C.F.R. § 73.3555(a)(1).
First, the Commission defended its use of numerical limits from those commenters who argued that there should be no limits at all on local radio station ownership. The Commission explained that radio is a closed-entry market, ie., all available radio frequency spectrum has been licensed. So the only way to develop market power is to buy existing stations. Thus, the Commission says, numerical limits are a reasonable way of preventing available capacity from being “locked up” in the hands of a few owners. Order ¶ 288.
The Commission then explained that the specific numerical limits that it currently employs are effective to this end. Citing economic literature, the Commission suggested that a market with five equal-sized competitors is a sufficiently competitive market. The existing limits allow for roughly five equal-sized firms in each market. (For example, the eight-station limit for markets of 45 or more stations allows for five eight-station combinations). Id. ¶ 289. The Commission concluded that the existing limits ensure sufficiently competitive local markets.76 The Commission also concluded that the existing limits are not overly restrictive based on data showing that the top-four stations in each metro market are thriving in terms of revenue and audience share. Id. ¶ 290.
As explained in Part II.B above, § 202(h) operates to extend the “reasoned analysis” requirement (which ordinarily applies only to an agency’s decision to promulgate new regulations or modify existing ones) to apply also to the Commission’s decision to retain existing regulations. Although we accept the Commission’s rationale for employing numerical limits (as opposed to other regulatory approaches such as a case-by-case analysis), we conclude that the Commission’s decision to retain the specific numerical limits is not supported by a reasoned analysis.
1. The Commission’s numerical limits approach is rational and in the public interest.
The Commission’s decision to retain a numerical limits approach to radio station ownership regulation is “in the public interest.” Without numerical limits, radio markets risk becoming “locked up” in the hands of a few owners (or even one owner) because all of the available radio frequency spectrum has been licensed — a high barrier to new market entrants. Order ¶288. Based on record evidence, the *432Commission justifiably concluded that numerical limits are necessary “to guard against consolidation ... and to ensure a market structure that fosters opportunities for new entry into radio broadcasting.” Id. ¶ 291. For example, a MOWG study found that, since the existing limits were imposed in 1996, the number of radio station owners declined by 34% even though the number of stations increased by 5.4%. George Williams & Scott Roberts, Radio Industry Review 2002: Trends in Ownership, Format, and Finance (MOWG Study No. 11) at 3 (Sept. 2002). Additionally, the record shows that today 10 parent companies — the largest of which, Clear Channel Communications, owns 1200 stations nationwide, or 10%' — dominate the radio industry and control about two-thirds of both listeners and radio revenues nationwide. Id. at 4. In contrast, prior to the 1996 Act’s deregulation, the largest nationwide radio station combinations had fewer than 65 stations each. Id.
Furthermore, the record shows how increased consolidation has increased station prices, which limits opportunities for new market entrants and as a result limits diversity in station ownership and output. UCC Comments at 18. Consolidation has also reduced the amount of locally produced radio content, as large group-owners often broadcast remotely from national offices instead of having local employees produce programming. Comments of Future of Music Coalition, MB Docket 02-277 at 13-14 (Nov. 20, 2002). The record contains examples of consolidated stations that eliminated local news production. UCC Comments at 37-38. The evidence thus supports the Commission’s conclusion that, by continuing to limit the consolidation of radio stations, numerical limits are “in the public interest” as required under § 202(h).
2. The Commission did not support its decision to retain the existing numerical limits with reasoned analysis.
Both the Citizen Petitioners and the Deregulatory Petitioners argue that the Commission’s decision to retain the existing numerical limits was arbitrary and capricious. Predictably, the Citizen Petitioners argue that the Commission should have tightened the existing limits, and the Deregulatory Petitioners argue that the Commission should have relaxed them. But both sides’ predominant argument is essentially the same: the numerical limits are not supported by. the Commission’s theory that they ensure five equal-sized competitors in most markets. While, as discussed above, substantial evidence supports the Commission’s decision to retain the numerical limits structure of its local radio ownership rule, we also agree with the Petitioners that the Order lacks a reasoned analysis for retaining these specific numerical limits. We thus remand for the Commission’s additional justification.
a. The Commission did not sufficiently justified “five equal-sized competitors” as the right benchmark.
The Commission relied on game theory to support its premise that five equal-sized competitors ensure that local markets are fragmented and structurally competitive. Order ¶ 289 n. 609 (citing Louis Phillips, Competition Policy: A Game Theory Perspective Ch. 2 (1995); Timothy F. Bresnahan & Peter C. Reiss, Entry and Competition in Concentrated Markets, 99 J. Pol. Econ. 997 (1991); Reinhard Selten, A Simple Model of Imperfect Competition Where Four Are Few and Six Are Many, 2 Int’l J. Game Theory 141 (1973)). The Citizen Petitioners and the Deregulatory Petitioners both dispute that these articles support the Commission’s selection of five equal-*433sized competitors as the appropriate benchmark. The Deregulatory Petitioners argue that the articles fall short because they do not rule out market structures other than equal-sized competitors (such as one large firm and many small ones) as equally competitive markets. This argument, unanswered by the Commission, warrants a response because (as discussed below) the record evidence supports neither actual nor potential existence of equal-sized competitors.
The Citizen Petitioners argue that the three articles the Commission cites are superseded and contradicted by the Department of Justice and Federal Trade Commission’s most recent re-write of the Merger Guidelines, under which a market with five equal-sized competitors is considered “highly concentrated.” See Merger Guidelines § 1.51(c) (a highly concentrated market has an HHI score higher than 1800). The Commission’s decision to rely on theory that conflicts with the Merger Guidelines is suspect because, in the same Order, the Commission relied on the Merger Guidelines to derive its new limits for local television station ownership. See Order ¶¶ 192-93. The Commission did not address this discrepancy in its Respondent’s Brief,77 and should do so on remand.
b. The Commission did not sufficiently justified that the existing numerical limits actually ensure that markets will have five equal-sized competitors.
Regardless whether five equal-sized competitors is the right benchmark for competition, the Commission did not sufficiently justified that five equal-sized competitors would emerge or actually have emerged under the numerical limits. It defies logic to assume that a combination of top-ranked stations is the competitive equal to a combination of low-ranked stations just because the two combinations have the same number of stations. The Commission .itself acknowledges that “radio station groups with similar numbers of radio stations [can] have vastly different levels of market power.” Order ¶ 290.
Furthermore, evidence shows that the existing numerical limits do not ensure five equal-sized competitors. According to the record, most markets are dominated by one or two large station owners.78 And the top-four station owners together control the lion’s share of the market.79 Even *434if these four station-owners were “equal-sized” (they are not), the HHI score of such a market would be 2500 (4 x (100 4)2), well above the Merger Guidelines’ 1800 threshold for highly concentrated markets.
The Commission does not explain why it could not take actual market share into account when deriving the numerical limits.80 Had it proffered the “market share is too fluid” rationale, we have already rejected that explanation in the context of the local television ownership rule and the Cross-Media Limits. We also note that the Commission has in the past extolled the value of audience share data for measuring diversity and competition in local radio markets.81 So the Commission’s reliance on the fiction of equal-sized competitors, as opposed to measuring their actual competitive power, is even more suspect in the context of the local radio rule.
For these reasons, the Commission’s numerical limits cannot rationally be derived from a “five equal-sized competitor” premise. We thus remand for the Commission to develop numerical limits that are supported by a rational analysis.
c. The Commission did not support its decision to retain the AlM/FM subcaps.
The Deregulatory Petitioners challenge the Commission’s decision to retain the AM/FM subcaps, which the Commission justified on the grounds that FM stations have technological and economic advantages over AM stations. Order ¶ 294. But the Deregulatory Petitioners point out, and we agree, that this does not explain why it is necessary to impose an AM subcap at all. The Commission does not respond in its brief to this particular criti-*435eism. Thus it should do so, or modify its approach, on remand.
VII. Conclusion
Though we affirm much of the Commission’s Order, we have identified several provisions in which the Commission falls short of its obligation to justify its decisions to retain, repeal, or modify its media ownership regulations with reasoned analysis. The Commission’s derivation of new Cross-Media Limits, and its modification of the numerical limits on both television and radio station ownership in local markets, all have the same essential flaw: an unjustified assumption that media outlets of the same type make an equal contribution to diversity and competition in local markets. We thus remand for the Commission to justify or modify its approach to setting numerical limits. We also remand for the Commission to reconsider or better explain its decision to repeal the FSSR.82 The stay currently in effect will continue pending our review of the Commission’s action on remand, over which this panel retains jurisdiction.
. For convenience we will use this designation throughout our opinion to refer, jointly or severally, to the petitioners and intervenor who raise anti-deregulatory challenges to the Order, including Prometheus Radio Project, Media Alliance, National Council of the Churches of Christ in the United States, Fairness and Accuracy in Reporting, Center for Digital Democracy, Consumer Union and Consumer Federation of America, Minority Media and Telecommunications Council (representing numerous trade, consumer, professional, and civic organizations concerned with telecommunications policy as it relates to racial minorities and women), and Office of Communication of the United Church of Christ (“UCC”) (intervenor).
The Network Affiliated Stations Alliance, representing the CBS Television Network Affiliates Association, the NBC Television Affiliates, and the ABC Television Affiliates, and Capitol Broadcasting Company, Inc. (interve-nor) also raised anti-deregulatory challenges to the national television ownership rule (see infra Parts I.F.5, III).
. Rather than identifying these petitioners and intervenors individually when discussing their positions throughout this opinion, we refer to them collectively as the "Deregulatory Petitioners.” They include: Clear Channel Communications, Inc.; Emmis Communications Corporation; Fox Entertainment Group, Inc.; Fox Television Stations, Inc.; Media General Inc.; National Association of Broadcasters; National Broadcasting Company, Inc.; Paxson Communications Corporation; Sinclair Broadcast Group; Telemundo Communications Group, Inc.; Tribune Company; Viacom Inc.; Belo Corporation (intervenor); Gannett Corporation (intervenor); Morris Communications Company (intervenor); Mill-creek Broadcasting LLC (intervenor); Nassau Broadcasting Holdings (intervenor); Nassau Broadcasting II, LLC (intervenor); Newspaper Association of America (intervenor); and Univision Communications, Inc. (intervenor).
. Our dissenting colleague asserts several times (indeed it is a principal theme) that we somehow substitute our own policy judgment for that of the Commission. Our response is simple. It is impossible to substitute our policy judgment for that of the Commission when we have no view on its policies save that it act with reason. The proof for this statement is that we do not reverse, but remand (and only in part).
Differences with our dissenting colleague touch only some of the many issues presented on appeal. He believes that the Commission's work, while in part flawed, comes close enough to merit our approval. Our response: not yet. He believes that any imperfections and failings by the Commission can be rectified at the next quadrennial review without the need for a court-ordered remand. Our response: to do so is to abdicate the role assigned to us (see our standard of review discussion at Part II infra), a role, incidentally, our colleague does not deny. What we do is not novel, as it reprises many of the same reasoned analysis concerns expressed by the D.C. Circuit Court of Appeals in Fox Television Stations v. FCC, 280 F.3d 1027 (D.C.Cir.2002), modified on reh'g, 293 F.3d 537 (D.C.Cir.2002), and Sinclair Broadcast Group Inc. v. FCC, 284 F.3d 148 (D.C.Cir.2002).
The bottom line: the Commission gets another chance to justify its actions. Once that occurs and the case returns to us, we may yet close the loop of agreement with our colleague.
. VHF (Very High Frequency) stations (chan-neis 2 to 13), which broadcast on frequencies *384between 30 and 300 MHz, can reach households up to 76 miles away. UHF (Ultra High Frequency) stations (channels higher than 13), which broadcast between 300 and 3000 MHz-reach only 44 miles, and, consequently, far fewer non-cable households than VHF stations. Order ¶ 586.
. These limits allowed an entity to own one television station (or two if permitted under the new local television rule) and (1) up to six radio stations in markets where at least 20 independent voices would remain post-merger, (2) up to four radio stations in markets where at least 10 independent voices would remain post-merger, or (3) one radio station notwithstanding the number of independent voices in the market. Id.
. The Commission also announced that it would propose modification of the newspaper/broadcast cross-ownership rule and the dual network rule (which allows a broadcast station to affiliate with more than one network except that stations may not affiliate with more than one of the four largest networks: ABC, CBS, Fox, and NBC, see 47 C.F.R. § 73.658(g)). The Commission also found that the local radio rule was necessary in the public interest, but announced that it would seek further comments on alternative methods for defining radio station markets. 1998 Biennial Regulatory Review ¶ 4.
. Two Commissioners (Jonathan S. Adelstein and Michael J. Copps) held and attended, between them, thirteen additional public forums, using their own office resources. Statement of Commissioner Michael J. Copps, Dis-sen ting, 18 F.C.C.R. 13,951, 13,956 (July 2, 2003).
. Commissioners Copps and Adelstein dissented.
. In the 1996 Act, Congress directed the Commission to revise the local radio ownership limits to provide that: (1) in a radio market with 45 or more commercial radio stations, a party may own up to 8 commercial radio stations, not more than 5 of which are in the same service (AM or FM); (2) in a radio market with between 30 and 44 commercial radio stations, a party may own up to 7 commercial radio stations, not more than 4 of which are in the same service; (3) in a radio market with between 15 and 29 commercial radio stations, a party may own up to 6 commercial stations, not more than 4 of which are in the same service; and (4) in a market with 14 or fewer commercial radio stations, a party may own up to 5 commercial radio stations, not more than 3 of which are in the same service, except that a party may not own more than 50 percent of the stations in that market. 1996 Act, § 202(b)(1), 110 Stat. at 110.
. No party to this case challenges the retention of this rule.
. The Deregulatory Petitioners raise an additional threshold argument that Fox and Sinclair provide the “law of the case” because the Order results in part from the D.C. Circuit Court's remand in those two decisions. Though we recognize that the Order must be consistent with these remand directives, the decisions themselves do not fit within the law of the case doctrine, under which a prior decision binds only future proceedings in the "same litigation.” See, e.g., Hamilton v. Leavy, 322 F.3d 776, 786-87 (3d Cir.2003). This case involves petitions for review of the Commission's comprehensive reexamination of a larger set of its broadcast ownership rules, in which a different set of parties participated, a different record was compiled, and different results were reached. So understood, the law of the case doctrine does not constrain our review here.
. As noted in Part I.H above, Congress has since replaced "biennially’' with "quadrennially" in this provision.
. More generally, the Commission cited several Supreme Court decisions interpreting "necessary” to mean "useful” or "convenient” or "appropriate.” Id. ¶ 15 n. 24 (citing, inter alia, McCulloch v. Maryland, 17 U.S. (4 Wheat.) 316, 413, 4 L.Ed. 579 (1819) (under the Constitution's Necessary and Proper Clause, U.S. Const. Art. I, § 8, cl. 18, "necessary” means "convenient” or "use-fill”)).
. In addressing available remedies, the Fox I Court interpreted "necessary” to mean that "a regulation should be retained only insofar as it is necessary in, not merely consonant with, the public interest.” 280 F.3d at 1050. This language was expressly excised in Fox II, 293 F.3d at 540.
. Additionally, the Commission itself impliedly incorporated the 2002 Biennial Regulatory Review's interpretation of “necessary” *394under § 11 when citing it in the "legal framework” section of the Order. Order ¶ 11 n. 15.
. Because § 202(h) omits the word "necessary” in the "repeal or modify” instruction, it is arguably even more amenable to the "appropriate” or "useful” connotation of “necessary" than § 11. But like the Commission, "we do not think that difference [between § 11 and § 202(h)] is significant given what we believe 'necessary in the public interest’ means.” 18 F.C.C.R. 4726, ¶ 14 n. 20.
. Anticipating the argument that consistency between the first and second sentences could still be achieved using the "indispensable” interpretation of "necessary” for both, we respond that requiring the Commission to repeal or modify any regulations it determined were not essential to the public interest would lead to incongruous results when compared to the plain public interest standard that governs the Commission’s authority to regulate the broadcast industry. See NCCB, 436 U.S. at 796, 98 S.Ct. 2096 (”[S]o long as the regulations are not an unreasonable means for seeking to achieve the [Commission's public interest] goals, they fall within [its] general rulemaking authority” under the Communications Act.). Just as the Commission and Céli-co observed in the § 11 context, the "repeal or modify” instruction would be meaningless because the Commission could repromulgate (in its rulemaking functions) under the lower "useful” standard any regulation it was forced to repeal or modify for not being "indispensable” under § 202(h).
. For example, in enacting a periodic review requirement for the Commission’s broadcast media ownership regulations, Congress gave no express indication that it intended to restrict the Commission’s rulemaking authority. See, e.g., Am. Hosp. Ass’n v. NLRB, 499 U.S. 606, 613, 111 S.Ct. 1539, 113 L.Ed.2d 675 (1991) (stating that “if Congress had intended to curtail in a particular area the broad rule-making authority [it has] granted!,] ... we would have suspected it to do so in language expressly describing an exception [to that authority]”)
. A purely textual inquiry is also of no avail to the Deregulatory Petitioners. As we noted above, § 202(h) does not use the word "necessary” in the second sentence to qualify the public interest standard that governs the Commission's “repeal or modify” instruction. Like the Commission (as we point out in note 16 supra), "we do not think that difference ... is significant given what we believe 'necessary in the public interest’ means.” 18 F.C.C.R. 4726, ¶ 14 n. 20. But even if "necessary in the public interest” under the first sentence meant "indispensable,” we still would not have to import that heightened standard to the second. Textually it is not there. If "necessary” appears twice in § 11 and in the first sentence of § 202(h), but is absent from its second sentence, logic favors deliberate omission by Congress over inadvertence.
. Although our dissenting colleague states that he differs "from the majority on the applicable standard of review,” we can discern no real disagreement between his formulation of the standard of review and ours. Rather, we see disagreement only over the question of whether the Commission’s Order survives the standard of review as we both have described it.
. The Deregulatory Petitioners stated that the Commission's rationale for retaining a national television ownership limit "raises troubling First Amendment questions.” Br. of Petitioners Fox, NBC, Telemundo, and Viacom ("Network Petitioners”) at 31. We recognize that a constitutional challenge would not necessarily be mooted by intervening legislation. But even if we interpreted this "troubling ... questions” language as raising a constitutional challenge to the 45% national audience reach limit, we see no reason to decide the constitutionality of Congress's 39% limit, as these same Petitioners subsequently argued that "any pending challenges to the June order [regarding the national television ownership rule] will be mooted” by the 2004 Consolidated Appropriations Act. Network Petitioners’ Letter Br. at 2 (Feb. 2, 2004).
. The Commission also repealed its radio/television cross-ownership rule in favor of the Cross-Media Limits, but no petitioner challenges this aspect of the Order.
. For example, Gannett Company reported that the "media integration” resulting from its newspaper/broadcast combination in Phoenix "improved efficiency, particularly in situations characterized by fast-breaking news such as the massive wildfires near Phoenix last year.” Both of Belo Corporation’s Dallas outlets "have been able to cover a wider range of stories through information sharing between the separate newspaper and television news staffs." Order ¶ 348.
. On one hand, the record contained examples of a newspaper-owner’s affiliations and biases influencing the news and editorial pages. But there were also examples of commonly owned newspapers expressing different viewpoints (such as that the Tribune Company's newspapers did not endorse the same candidate in the 2000 presidential election). The MOWG submitted a statistical study of 10 newspaper-television combinations; in half of them the newspaper's overall "slant” was noticeably different from the broadcast television's "slant” on the news. See David Pritchard, Viewpoint Diversity in Cross-Owned Newspapers and Television Stations: A Study of News Coverage of the 2000 Presidential Campaign (MOWG Study No. 2) (Sept. 2002). But the Pritchard Study was the subject of much criticism for its flawed methodology (no control group of independent media outlets for comparison) and narrow scope (it only looked at one factor in judging "slant”). Order ¶ 362.
. See, e.g., Joel Waldfogel, Consumer Substi-tutability Among Media (MOWG Study No. 3) (Sept.2002) (suggesting that consumers view Internet news sources as a substitute for daily newspapers and broadcast news).
. Indeed, ample evidence supported its conclusion that ownership can influence viewpoint. See Order ¶¶ 24-25; Comments of the UCC et al., MB Docket No. 02-277 at 4 (Jan. 2, 2003) {“UCC Comments’’) (citing a Pew Research Center study of television journalists and executives that found nearly one quarter of journalists purposefully avoid newsworthy stories and nearly as many soften the tone of stories to benefit the interest of their news organizations); id. at 5 (citing a 2002 study finding that outlets included more references to their own products and services and treated those items more favorably than others, thus exhibiting a "synergy bias”); Comments of Consumer Federation of America, MB Docket 02-277 at 41 (Jan. 2, 2003) (citing a 2002 study finding that election information on news pages was slanted in favor of candidate endorsed on the editorial page); id. at 44 (citing a 2001 survey of news directors finding that media owners and sponsors pressure reporters to slant the news).
. We also note that the Commission provided for a waiver of the Cross-Media Limits upon a demonstration "that an otherwise prohibited combination would, in fact, enhance the quality and quantity of broadcast news available in the market.” Order ¶ 481.
. Of course, the HHI formula also works where the components of a market are not of equal size. In a market where one company controls 50% of the market share, two control 20%, and another controls 10%, the HHI formula is 502 + 202 + 202 + 102 = 3400.
. The Commission calculated Diversity Index scores for all markets with five or fewer television stations, all markets with 15 and 20 television stations, and ten randomly selected markets with between six and ten stations. For sample market Diversity Index calculations, see Order app. C.
.Throughout this discussion of the Diversity Index, we adopt the Commission's convention and report the decimals of intermediate prod*404ucts and quotients (such as 4.3 here) rounded to the nearest tenth. Final products (such as 13.0 reported in the next sentence) are derived from the unrounded intermediate product and then rounded to the nearest tenth.
. The dissent suggests that the MOWG Studies, "although not perfect ..are a significant improvement over the study in Sinclair” that the D.C. Circuit Court determined did not support the Commission’s decision to limit the definition of "voices” in its local television rule. But this comparison is not relevant because, as we point out, the Commission inconsistently applied the results of its analysis.
. The follow-up question for the respondents who reported using the Internet as a news source asked which sites they had used in the past seven days as a source of local or national current affairs, so their responses to this question are not helpful in determining whether the Internet .is a source of independent local news. To the extent that wé can glean anything from this question — unlikely, as the predominant response was "other” (34.9%) — the most popular websites identified were those of national cable news channels, MSN.com (22.4%) and CNN.com (19.1%), indicating that most people had those types of websites in mind when they reported using the Internet for news. (The other responses were: Yahoo.com (17.9%), MSNBC.com (5.9%), NewYorkTimes.com (5.1%), Fox-News.com (2.2%), USAToday.com (1.8%), AB-*406CNews.com (1.8%), CBSNews.com (1.7%), Netscape.com (1.7%), Excite.com (1.3%), Iwon.com (1.2%), AT&T.net (1.1%), WallS-treetJournal.com (1.0%), none (7.3%), don’t know (3.5%), and refuse (0.5%)).
.The dissent acknowledges the flaws in the MOWG Study No. 8 survey, but suggests that because the survey "can be fine-tuned on the next round of reviews” it should not be the basis for remand. But we decline to abdicate our obligation to provide meaningful judicial review. Just as the Commission may not rely on a "wait-and-see” approach to rulemaking, Sinclair, 284 F.3d at 164, neither can we. We review an agency’s decision for whether it is rationally derived from the record evidence — not whether the agency may (or may not) "fine tune” the record at the next man- ' dated review. Furthermore, just as we may not supply a reasoned basis for the agency’s action that the agency itself has not given, State Farm, 463 U.S. at 43, 103 S.Ct. 2856, we also do not supply evidence to an agency in the hopes that the agency will include that evidence in a future study.
. Moreover, the Drudge Report is an "aggre-gator” of news stories from other news outlets’ websites and, as such, is not itself normally a "source” of news, national or local.
. The Order also speaks of "news and public affairs programming” or "local news and current affairs information” when referring to measures of viewpoint diversity. E.g., Order IN 394, 406, 409.
. The dissent also takes notice of the Independent Media Center websites such as philly-imc.com. We agree that this is a good example of independent local news on the Internet. But the IMC provides local links in only eight markets in the United States. The Commission does not even mention these websites in its analysis of the Internet, let alone argue that IMC websites in eight local markets will offset the viewpoint diversity lost when newspapers and broadcast stations are allowed to consolidate. And even if we believed that to be the case, it is not our role to insert our own analysis to substitute for that which is missing from the agency's rationale. State Farm, 463 U.S. at 43, 103 S.Ct. 2856 ("[W]e *407may not supply a reasoned basis for the agency’s action that the agency itself has not given.”).
. See Brief of Petitioner Media General, Inc., at 55 (noting "each media outlet in a community is actually a platform for the expression of many viewpoints”).
. The dissent views these flaws, effectively, as "harmless error” because the Commission called the Diversity Index a "useful tool” that "informs, but does not replace, our judgment in establishing rules of general applicability that determine where we should draw lines between diverse and concentrated markets.” Order ¶ 391. But nowhere in the Order, in its briefs, or in its oral argument did the Commission identify any consideration other than the Diversity Index as having influenced the formulation of the Cross-Media Limits. Rather, the Cross-Media Limits prospectively ban certain combinations in specific markets, and allow others, based on nothing but the relative increases those combinations would have on the average Diversity Index scores of markets of that size.
. Source: Audit Bureau of Circulations. See http://www.accessabc.com/.
. The Commission did not report a Diversity Index increase for those scenarios that would be precluded by the modified local television ownership rule, see generally Part v. infra, which would prohibit television station duopolies in markets with fewer than five stations.
. The dissent suggests that the Commission provides sufficient justification in the Order when it explains why the newspaper + 2 TV stations combination is not allowed in midsized markets but the newspaper + 1 TV station + 50% of allowed radio stations combination is allowed. Order ¶ 467 (suggesting that a newspaper will benefit more from the consolidation with its first-acquired TV station than with subsequently acquired stations). But this does not address why the newspaper + 1 TV station + 50% allowed radio stations combination was permitted when its Diversity Index score increases were overall much greater than the Diversity Index score increases for other allowed combinations. In other words, the Commission may have proffered an explanation for why "F” should be treated differently from seemingly similar "D,” but it does not explain why "D” should be treated similarly from seemingly different "A,” "B,” and "C.”
. The Commission also argues that the Diversity Index was "simply an analytical tool" for measuring diversity. Resp’t Br. at 90-91. But as we noted before, see supra note 38, it offers no suggestion that anything other than the Diversity Index was used to formulate the Cross-Media Limits.
. The Commission considers two field-strength contours, Grade A and Grade B, as indicators of a station’s approximate extent of coverage over average terrain in the absence of interference. Grade B contours measure a weaker signal than Grade A, and thus have a wider coverage area. 47 C.F.R. § 73.683.
. Under the new rule, the top-four restriction may be waived on a case-by-case basis. In deciding whether to grant a waiver, the Commission may consider such factors as ratings information demonstrating the competitive effect of the merger, the proposed merger's effect on the stations' ability to transition to a digital signal, and the proposed merger’s effect on localism and viewpoint diversity. Order ¶¶ 227-30. The top-four restriction waiver, however, is only available in markets of 11 or fewer stations.
. This finding is consistent with the Commission's 1999 decision to relax the local television ownership rule because local ownership combinations were likely to yield efficiencies that “can in turn lead to cost savings, which can lead to programming and other service benefits that enhance the public interest.” Order ¶ 155 (citing 1999 Television Rule Review, 14 F.C.C.R. 12,903, ¶ 34).
. The Commission found that 85% of top-four stations offer local news programs, as compared to 19% of stations outside the top four. Order ¶ 198 (citing Bruce M. Owen et al., News and Public Affairs Programming Offered by the Four Top-Ranked Versus Lower-Ranked Television Stations, included in Comments of Fox Entertainment Group, Inc., et al., MB Docket 02-277 (Jan. 2, 2003)).
. As explained more fully in the next sub-part, the Commission justified its numerical limits on the belief that they would ensure six equal-sized competitors in- most markets. But the Commission departed from this rationale when it allowed duopolies in markets with five to eleven stations. Order ¶ 201. The Commission explained that stations in those small and mid-sized markets are experiencing greater competitive difficulty than stations in large markets. Id. (citing data on the comparative profitability of stations in markets of various sizes). Thus, the Commission determined, it was necessary to allow some additional concentration in those markets (relative to that allowed in the larger markets) so that small and mid-sized market stations would realize the efficiency benefits of consolidation.
. The Deregulatory Petitioners challenge the waiver provision as well, suggesting that the Commission’s decision to limit its availability to markets with fewer than 12 stations is an arbitrary and capricious number “plucked out of thin air.” See Sinclair, 284 F.3d at 162. But the Commission explained that its decision to draw the line at markets with fewer than 12 stations was based on its determination that it was in those markets that "the economics of broadcast television justify relatively greater levels of station consolidation.” Order ¶ 227. Moreover, it is consistent with the Commission’s overall approach of maintaining a balance between ensuring that small-market stations can reap the benefits of consolidation while protecting the public’s interest in viewpoint diversity. Thus we defer to the Commission's expertise and affirm this particular line-drawing.
We also note that the modified rule does not preclude a top-four restriction waiver in a market with more than 12 stations, as the Commission expressly acknowledged its duty to give a "hard look ... to waiver requests.” Id. ¶ 85; see also 47 C.F.R. § 1.3 (authorizing waiver of Commission rules "for good cause shown”).
. In all but 18 of the markets ranked 76 to 201, the combined market share of the third and fourth-ranked station is less than that of the top-ranked station. Ex Parte Communication of National Association of Broadcasters, MB Docket No. 02-277 at 4 (May 22, 2003).
. In the largest 150 markets, the fourth-ranked station trailed the third-ranked station by 34% in audience share and 26% in revenue share. Bear, Stearns & Co., Inc., Duopoly Relief Needed-4th Ranked Stations Significantly Trail 3rd Ranked Stations (May 29, 2003).
. While the likelihood that a third and fourth-ranked station combination would produce a new largest station was lower in the mid and small-sized markets, the market share of those markets' top-four stations combined was much higher than in the large markets. Id. So while the top-four restriction does not operate to protect the small and midsized markets from as many "new largest station” mergers, it furthers diversity by ensuring there are at least four independent voices in those markets.
. The Deregulatory Petitioners also challenge the Commission’s six equal-sized competitor benchmark as inconsistent with its five equal-sized competitor benchmark in the local radio rule. Order ¶ 289. No reason exists, however, for the Commission's local television ownership limits to mirror precisely its *419local radio ownership limits, particularly given that there are generally more radio stations than television stations in a given market.
. The Commission also defends its rule against charges that it is duplicative of antitrust authority by pointing out that it takes audience preferences into account, unlike antitrust regulators who focus on prices:
. The Commission deliberately did not select the Merger Guidelines’ 1000 threshold for moderately concentrated markets out of recognition that television stations receive competitive pressure not only from each other but from cable networks as well.
We reject the Deregulatory Petitioners' contention that the Commission's consideration of the competitive effect of cable for this purpose is inconsistent with its refusal to define the relevant competitive market to include cable networks. The Commission explained that cable networks offer "almost exclusively ... national or broadly defined regional programming,” and thus profit-maximizing decisions reflect national, rather than local, markets. Order ¶ 191. On this ground, the Commission justified its decision not to accept that television and cable compete in the same market. As a matter of degree, however, the Commission recognized that cable networks exerted some competitive pressure on local networks, and thus selected a higher benchmark (1800) them it otherwise might have.
.The dissent says that the Commission selected the HHI score of 1800 as a mere "starting point” for its analysis. Dissent Part IV.B. Indeed', the Commission cautioned against "strict, overly simplistic application of the DOJ/FTC Merger Guidelines.” Order ¶ 192. But the triopoly rule was justified by the six *420equal-sized competitor rationale, which was in turn derived from the Merger Guidelines' 1800 benchmark. No other rationale was provided. Surely, then, it is proper judicial review to call the Commission on its failure to adhere to the 1800 benchmark it selected.
. The proponent of this argument is the Minority Media and Telecommunications Council ("MMTC”), which had participated in the briefing as an intervenor party. On April 7, 2004, MMTC withdrew its petition to the Commission for reconsideration on the same grounds, thus eliminating the jurisdictional bar that otherwise would have precluded our review. See West Penn Power Co. v. EPA, 860 F.2d 581, 587 (3d Cir.1988). We then accepted MMTC’s petition for review because its pending reconsideration petition tolled the time for filing a petition for judicial review. See L.A. SMSA Ltd. P’ship v. FCC, 70 F.3d 1358, 1359 (D.C.Cir.1995). MMTC's new status as a petitioner rather than an intervenor negates the Commission’s initial concern that MMTC had impermissibly expanded the scope of issues on review by raising an argument that was not addressed in any of the petitioners' briefs.
. We fail to see the logic in this proffered rationale. Even if it were true that same-market efficiencies will always lead to a duopoly absent unusual circumstances, it does not follow, without additional proof or explanation, that (1) marketing the station outside the market is a meaningless burden or that (2) the Commission should not retain the FSSR to make that "circumstance” less "unusual.”
. Repealing its only regulatory provision that promoted minority television station ownership without considering the repeal's effect on minority ownership is also inconsistent with the Commission's obligation to make the broadcast spectrum available to all people "without discrimination on the basis of race.” 47 U.S.C. § 151.
. We also note that the Commission deferred consideration of the MMTC's other proposals for advancing minority and disadvantaged businesses and for promoting diversity in broadcasting. See Order ¶¶ 49-50 (promising to issue a Notice of Proposed Rulemaking to address the MMTC’s thirteen specific proposals); id. ¶ 52 (deferring consideration of the MMTC’s "Transaction Nondiscrimination” proposal pending recommendations from its Advisory Committee for Diversity). The Commission’s rulemaking process in response to our remand order should address these proposals at the same time.
.Specifically, the numerical limits provided that:
(1) In markets with fewer than 15 radio stations, a single licensee will be permitted to own up to three stations, no more than two of which are in the same service, provided that the owned stations represent less than 50 percent of the stations in the market. Common ownership of one AM/FM combination will continue to be allowed in any event.
(2) In markets with 15 to 29 radio stations, a single licensee will be permitted to own up to two AM stations and two FM stations, provided that the combined audience share of the stations does not exceed 25 percent.
*422(3) In markets with 30 to 39 radio stations, a single licensee will be permitted to own up to three AM stations and two FM stations, provided that the combined audience share of the stations does not exceed 25 percent.
(4) In markets with 40 or more radio stations, a single licensee may own up to three AM stations and three FM stations, provided that the combined audience share of the stations does not exceed 25 percent.
57 Fed. Reg. at 18090.
. Congress's 1996 Act-directed limits are set out in full at Part I.F.2 supra.
. The 1996 Act did not define radio markets by any particular methodology.
. In addition, the Commission noted that the Department of Justice treats Arbitron Metros as the relevant market for antitrust purposes. Order ¶ 276.
. The Commission included noncommercial stations upon finding that they "exert[] competitive pressure on all other radio stations in the market seeking to attract the attention of the same body of potential listeners.” Order ¶ 295.
. In 33 markets of various sizes, investment banking firm Bear Stearns found that there would be an average 47.5% fewer stations under the Commission's adopted approach (that includes noncommercial stations and uses Arbitron Metro markets) than under the existing approach (that excludes noncommercial stations and employs the contour-overlap methodology). In only 6 of the 33 markets studied by Bear Stearns was there an increase in market size under the adopted approach. See Bear, Stearns and Co., A Defining Moment in Radio?, MB Docket No. 02-277 (May 12, 2003).
. We also reject the Citizen Petitioners' contention that Commission failed to provide adequate notice of its decision to include noncommercial stations in its new market definition. The Notice stated that the Commission would engage in a "comprehensive review” of its media ownership rules, including, specifically, whether its competition analysis should focus on "competition for viewers/listeners,” "competition for programming,” and "competition for advertising.” 17 F.C.C.R. 18,503, ¶¶ 50, 52, 57. This language adequately apprised the public that the Commission might find, as it did, that noncommercial stations compete with commercial stations for listeners.
.Though the Commission restricted the transfer of existing combinations rendered noncompliant by all three local rules (Cross-Media Limits, local radio ownership rule, and local television ownership rule), we consider the Deregulatory Petitioners' challenges to the transfer restrictions issue within the context of the local radio rule because the modifications to the local radio rule (i.e., the change to Arbitron Metro markets) are more likely than the deregulatory modifications to the other two local rules to render existing combinations noncompliant.
. An eligible entity is a business with $6 million or less in annual revenue. Order ¶ 489.
. The Deregulatory Petitioners argue that the transfer restriction, as an anti-deregula-tory modification, is per se invalid under § 202(h). As we explained in Part II, we do not read § 202(h) to impose such a stringent standard for rulemaking. For our review purposes, any modification, whether d'eregu-latory or regulatory, that is "in the public interest” survives threshold scrutiny under § 202(h).
.We also reject the Citizen Petitioners' contention that the Commission should have chosen "socially and economically disadvantaged businesses” (SDBs) as the waiver-eligible class instead of Small Business Administration-defined small businesses. It noted that small businesses often include women and minority-owned stations. Order ¶ 488. The Commission noted that, because of pending legislation, the definition of SDBs is currently too uncertain to be the basis of its regulation. Id. ¶ 488 n. 1042. We anticipate, however, that by the next quadrennial review the Commission will have the benefit of a stable definition of SDBs, as well as several years of implementation experience, to help it reevaluate whether an SDB-based waiver will better promote the Commission’s diversity objectives.
. It is perhaps telling that the Deregulatory Petitioners do not argue that "questions” raised by this aspect of the Order rise to the level of an actual constitutional violation. Br. of Pet’r Clear Channel at 49.
. The Fifth Amendment provides, "No person shall ... be deprived of life, liberty, or property, without due process of law; nor shall private property be taken for public use, without just compensation.” U.S. Const. Amend. V.
. In this way, the Commission found that JSAs were no different from their Local Marketing Agreement counterparts under the local television rule, which are attributable. Id. ¶ 318.
. Again we reject the Deregulatory Petitioners' arguments that, as an anti-deregulatoiy modification, this aspect of the Order is a per se violation of § 202(h). See supra Part II.B.
. Again it is telling that the Deregulatory Petitioners do not argue that ''concerns" raised by this aspect of the Order rise to the level of an actual constitutional violation. Br. of Pet'r Clear Channel at 58.
. The Commission also explained why the numerical limits depart from the five equal-sized competitor rationale. Of its decision not to adopt a 10-station limit for markets with 50 or more stations, the Commission said that it found no evidence that additional consolidation would increase efficiencies in those markets. Furthermore, the Commission said that the largest markets usually have many small, low-power stations, which make those markets appear more competitive than they actually are. A numerical limit that ensures even more than five equal-sized competitors in these largest markets provides a competition-protecting cushion that compensates for this. Order ¶291. The Commission also realized that its numerical limits for the smaller market tiers also depart from the five equal-sized competitor rationale. It maintained that, in small markets, “greater levels of concentration may be needed to ensure the potential for viability of radio stations.” Id. ¶ 292.
As explained below, we believe that the numerical limits lack a rational underpinning in the "five equal-sized competitor” theory the Commission advanced. But because we remand for further justification by the Commission, we do not address Petitioners’ arguments regarding the Commission’s departure from the five equal-sized competitor rationale.
. The Commission argued instead that we should defer to its line-drawing discretion. But in order to warrant judicial deference, an agency’s line-drawing decision must be justified by a reasonable explanation and cannot run counter to the evidence before it. Sinclair, 284 F.3d at 162; see also NCCB, 436 U.S. at 814-15, 98 S.Ct. 2096. On this aspect of the Order, the Commission’s decision is neither.
. The Local Radio Market Concentration Study, submitted with comments of the UCC to MM Docket No. 01-317 (Mar. 26, 2002), supports this conclusion. The UCC studied 33 local radio markets (using 2001 data from the BIA Financial Network) of various sizes and reported each market's four largest station owners’ local commercial share. In most, the largest and second-largest station owners had market shares considerably higher than the next largest owners. For example, in San Francisco the top two station owners controlled, respectively, 26.8% and 20% of the local commercial share; the third and fourth-largest owners controlled 13.4% and 12.1%, respectively. In Atlanta, the largest station owner had a 30.8% market share while the owners of the three next-largest stations had between 9% and 17% each. Medium and small markets were also dominated by one or two owners. The four largest station owners in Jacksonville, Florida owned 45.5%, 36.3%, 9.1%, and 2.5%. The four (only) station owners in Fargo, North Dakota owned 50.8%, 40.2%, 7.0%, and 2.0%.
.See id. In all but one of the 11 of the 50 largest markets analyzed by the UCC, the four largest firms controlled between 64 and 82% *434of the market share. In its 15 mid-sized markets, four largest firms controlled between 86 and 100% of the market share. And in the seven small markets it studied, the four largest firms controlled between 92 and 100% of the market share.
. The dissent suggests that this conclusion is undermined by two paragraphs in the Order, ¶¶ 300-01, in which the Commission justified its decision not to reinstate an audience share cap on mergers. But the flaw we find in the Commission’s analysis has nothing to do with its conclusion that an audience share cap would discourage radio stations from earning market share by investing in quality programming. Order ¶ 300. Our concern is, rather, that the Commission sought to justify its numerical limits by suggesting that they would allow a sufficient number of equal-sized competitors (five) to allow for a competitive market, when in fact there is no evidence that the resulting competitors would be anything approaching “equal-sized.”
. In the context of its decision in its 1992 rulemaking to retain a 25% audience share benchmark at which an acquisition will raise a prima facie concern that the transaction will lead to undue local concentration, the Commission said:
We recognize that there are some limitations to relying exclusively on market share data to weigh concentration in the local radio marketplace, as pointed out by petitioners. However, to the extent petitioners argue that ratings data are inherently unsuitable for purposes of analyzing local concentration, we disagree. We believe that audience share information will be useful in helping to measure diversity and competition. Specifically, audience share data can identify the most dominant stations in a market. And, compared to limits based solely on the number of stations involved, use of audience share is a means of accounting for the variety of types of stations — clear channel, regional, daytime, low and high power — that exist in various markets. Moreover, by continuing to consider this factor as part of our ownership limits, our rules may provide an incentive for stronger, successful stations to invest in other local stations with generally low audience shares — an outcome that is consistent with the purposes of this proceeding.
Revision of Radio Rules and Policies, 7 F.C.C.R. 6387, ¶ 47.
. On remand the Commission should also consider MMTC's proposals for enhancing ownership opportunities for women and minorities, which the Commission had deferred for future consideration.