Introduction and Summary of Argument

The Sherman Act is the “Magna Carta of free enterprise.” Verizon Commcn’s Inc. v. Law Offs. of Curtis V. Trinko, LLP, 540 U.S. 398, 415 (2004) (citation omitted). It directs itself “not against conduct which is competitive, even severely so, but against conduct which unfairly tends to destroy competition itself.” Spectrum Sports, Inc. v. McQuillan, 506 U.S. 447, 458 (1993). And it does so not to protect corporate or private interests, but from concern for consumer welfare and the public interest. Id. The goal of antitrust law thus “is not to redress losers of legitimate competition; [i]t is to proscribe actions with anticompetitive effect.” Apartments Nationwide, Inc. v. Harmon Publ’g Co., 78 F.3d 584 (6th Cir. 1996) (table); Omega Env’t, Inc. v. Gilbarco, Inc., 127 F.3d 1157, 1163 (9th Cir. 1997) (“[T]he antitrust laws were not designed to equip [a] competitor with [its rival’s] legitimate competitive advantage.”).

For almost four decades, CoStar Group, Inc. and CoStar Realty Information, Inc. (“CoStar”) have provided commercial real estate
(“CRE”) services, including CRE listings and auction services. Commercial Real Estate Exchange Inc. (“CREXi”), launched almost a decade ago, is attempting to build its own CRE platform. CoStar filed suit against CREXi in September 2020, alleging that CREXi “harvests content, including broker directories, from CoStar’s subscription database without authorization by using passwords issued to other companies.” See Dist. Ct. Dkt. 1. In response, CREXi filed eight antitrust counterclaims for violations of the Sherman Act (seven claims) and the Cartwright Act (one claim). Dist. Ct. Dkt. 146. The district court dismissed them all.

In asking this Court to reverse the district court’s decision, CREXi and its amicus make three critical errors. First, CREXi and the FTC try to recast the court’s analysis as incorrectly applying a “refusal-to-deal framework.” Doc. 24 (“FTC Br.”) 10; Doc. 21 (“CREXi Br.”) 32. But the district court did not apply any such framework. Nor did it borrow any of the elements this Court has found must be pleaded in refusal to deal cases. See FTC v. Qualcomm Inc., 969 F.3d 974, 994–95 (9th Cir. 2020). Instead, the court did what courts do when considering antitrust claims—analyze contractual language and market realities in light of the bedrock antitrust principle that “a business generally has the right to refuse to deal with its competitors.” Dist. Ct. Dkt. 340 (“Op.”) 3. There is nothing improper about analyzing antitrust claims against the backdrop of this (and other) “traditional antitrust principles.” Trinko, 540 U.S. at 411.

Second, CREXi and the FTC argue that CoStar’s contractual provisions with brokers are “de facto” exclusivity provisions that violate the Sherman Act. CREXi Br. 62–64; FTC Br. 17–19. Yet both fail to acknowledge that this Court has never “explicitly recognized a ‘de facto’ exclusive dealing theory.” Aerotec Int’l, Inc. v. Honeywell Int’l, Inc., 836 F.3d 1171, 1182 (9th Cir. 2016). A careful examination of this theory reveals that it lacks a doctrinal foundation, and that this Court’s cases, historical context, and administrability concerns all counsel strongly against recognizing this theory. In any event, even if this were a viable theory, it is unavailable to CREXi here because CoStar’s express contractual terms plainly do not “substantially foreclose[]” brokers from dealing with CREXi. Id.

Third, CREXi and the FTC both urge this Court to hold that allegations of supracompetitive prices alone are enough to adequately allege direct evidence of market power, and thus that the test applied in Rebel Oil Co., Inc. v. Atl. Richfield Co., 51 F.3d 1421, 1433 (9th Cir. 1995), is “wrong as a matter of law.” CREXi Br. 38. Not so. Direct evidence of market power is “only rarely available.” United States v. Microsoft Corp., 253 F.3d 34, 51 (D.C. Cir. 2001) (en banc) (per curiam). And as the Supreme Court has made clear, market power is “the ability to raise prices profitably by restricting output. Ohio v. Am. Express Co., 138 S. Ct. 2274, 2288 (2018) (quoting Areeda & Hovenkamp § 5.01) (emphasis in original). “[H]igh price alone” thus is not enough to infer market power. Coal. for ICANN Transparency, Inc. v. VeriSign, Inc., 611 F.3d 495, 503 (9th Cir. 2010); see Blue Cross & Blue Shield United of Wis. v. Marshfield Clinic, 65 F.3d 1406, 1412 (7th Cir. 1995) (Posner, J.). A plaintiff seeking to present direct evidence of market power needs to show more than prices above a competitive level. It must show “evidence of restricted output and supracompetitive prices.” Rebel Oil, 51 F.3d at 1434 (emphasis added); Forsyth v. Humana, Inc., 114 F.3d 1467, 1476 (9th Cir. 1997) (“With no accompanying showing of restricted output,” “hig[h] prices” and “high profits … fail[ ] to present direct evidence of market power.”), overruled on other grounds by Lacey v. Maricopa Cnty, 693 F.3d 896 (9th Cir. 2012).

The arguments pushed by CREXi and the FTC, if accepted, would open the floodgates to baseless antitrust suits. Recognizing claims based on the de facto exclusive dealing theory would allow a competitor to transform its rival’s plainly nonexclusive contractual language into exclusive dealing provisions and drag its rival into expensive, protracted discovery based on speculative allegations about third-party conduct. Indeed, there is no end to what a struggling competitor could do with such an amorphous doctrine. So too, permitting a party to establish direct evidence of market power through allegations of supracompetitive prices alone would contravene binding authority and bedrock antitrust principles.

In rejecting these arguments below, the district court properly concluded that CREXi failed to meet its pleading burden for its antitrust
counterclaims. This Court should affirm.

Read the full brief here.

ICLE Director of Law & Economics Programs Gus Hurwitz was a guest on The Cyberlaw Podcast, where he discussed the U.S. Supreme Court’s Murthy v. Missouri free speech case and a unanimous decision by the court on when a public official may use a platform’s tools to suppress critics posting on his or her social-media page. Other topics included AI deepfakes, the congressional bill to force the divestment of TikTok, and the Federal Trade Commission’s lawsuit against Meta. Audio of the full episode is embedded below.

INTEREST OF AMICI CURIAE

Amici Curiae are 24 former government officials, former judges, and intellectual property scholars who have developed copyright law and policy, researched and written about copyright law, or both. They are concerned about ensuring that copyright law continues to secure both the rights of authors and publishers in creating and disseminating their works and the rights of the public in accessing these works. It is vital for this Court to maintain this balance between creators and the public set forth in the constitutional authorization to Congress to create the copyright laws. Amici have no stake in the parties or in the outcome of the case. The names and affiliations of the members of the Amici are set forth in Addendum A below.[1]

SUMMARY OF ARGUMENT

Copyright fulfills its constitutional purpose to incentivize the creation and dissemination of new works by securing to creators the exclusive rights of reproduction and distribution. 17 U.S.C. § 106. Congress narrowly tailored the exceptions to these rights to avoid undermining the balanced system envisioned by the Framers. See 17 U.S.C. §§ 107–22. As James Madison recognized, the “public good fully coincides . . . with the claims of individuals” in the protection of copyright. The Federalist NO. 43, at 271–72 (James Madison) (Clinton Rossiter ed., 1961). Internet Archive (“IA”) and its amici wrongly frame copyright’s balance of interests as between the incentive to create, on the one hand, and the public good, on the other hand. That is not the balance that copyright envisions.

IA’s position also ignores the key role that publishers serve in the incentives copyright offers to authors and other creators. Few authors, no matter how intellectually driven, will continue to perfect their craft if the economic rewards are insufficient to meet their basic needs. As the Supreme Court observed, “copyright law celebrates the profit motive, recognizing that the incentive to profit from the exploitation of copyrights will redound to the public benefit by resulting in the proliferation of knowledge.” Eldred v. Ashcroft, 537 U.S. 186, 212 n.18 (2003) (quoting Am. Geophysical Union v. Texaco Inc., 802 F. Supp. 1, 27 (S.D.N.Y. 1992)). Accordingly, the Supreme Court and Congress have long recognized that copyright secures the fruits of intermediaries’ labors in their innovative development of distribution mechanisms of authors’ works. Copyright does not judge the value of a book by its cover price. Rather, core copyright policy recognizes that the profit motive drives the willingness ex ante to invest time and resources in creating both copyrighted works and the means to distribute them. In sum, commercialization is fundamental to a functioning copyright system that achieves its constitutional purpose.

IA’s unauthorized reproduction and duplication of complete works runs roughshod over this framework. Its concept of controlled digital lending (CDL) does not fall into any exception—certainly not any conception of fair use recognized by the courts or considered by Congress—and thus violates copyright owners’ exclusive rights. Expanding the fair use doctrine to immunize IA’s wholesale copying would upend Congress’s carefully-considered, repeated rejections of similar proposals.

Hoping to excuse its disregard for copyright law, IA and its amici attempt to turn the fair use analysis on its head. They acknowledge that the first sale exception does not permit CDL, as this Court made clear in Capitol Records, LLC v. ReDigi Inc., 910 F.3d 649 (2d Cir. 2018).[2] They also are aware that courts consistently have rejected variations on the argument that wholesale copying, despite a format shift, is permissible under fair use.[3] Nevertheless, IA and its amici ask this Court, for the first time in history, to create a first sale-style exemption within the fair use analysis. CDL is not the natural evolution of libraries in the digital age; rather, like Frankenstein’s monster, it is an abomination composed of disparate parts of copyright doctrine. If endorsed by this Court, it would undermine the constitutional foundation of copyright jurisprudence and the separation of powers.

The parties and other amici address the specific legal doctrines, as well as the technical and commercial context in which these doctrinal requirements apply in this case, and thus Amici provide additional information on the nature and function of copyright that should inform this Court’s analysis and decision.

First, although IA and its amici argue that there are public benefits to the copying in which IA has engaged that support a finding that CDL is fair use, their arguments ignore that copyright itself promotes the public good and the inevitable harms that would result if copyright owners were unable to enforce their rights against the wholesale, digital distribution of their works by IA.

Second, IA’s assertion of the existence of a so-called “digital first sale” doctrine—a principle that, unlike the actual first sale statute, would permit the reproduction, as well as the distribution, of copyrighted works—is in direct conflict with Congress and the Copyright Office’s repeated study (and rejection) of similar proposals. Physical and digital copies simply are different, and it is not an accident that first sale applies only to the distribution of physical copies. Ignoring decades of research and debate, IA pretends instead that Congress has somehow overlooked digital first sale, yet left it open to the courts to engage in policymaking by shoehorning it into the fair use doctrine. By doing so, IA seeks to thwart the democratic process to gain in the courts what CDL’s proponents have not been able to get from Congress.

Third, given that there is no statutory support for CDL, most libraries offer their patrons access to digital works by entering into licensing agreements with authors and their publishers. Although a minority of libraries have participated in IA’s CDL practice, and a few have filed amicus briefs in support of IA in this Court, the vast majority of libraries steer clear because they recognize that wholesale copying and distribution deters the creation of new works. As author Sandra Cisneros understands: “Real libraries do not do what Internet Archive does.” A-250 (Cisneros Decl.) ¶12. There are innumerable ways of accessing books, none of which require authors and publishers to live in a world where their books are illegally distributed for free.

No court has ever found that reproducing and giving away entire works—en masse, without permission, and without additional comment, criticism, or justification—constitutes fair use. IA’s CDL theory is a fantasy divorced from the Constitution, the laws enacted by Congress, and the longstanding policies that have informed copyright jurisprudence. This Court should reject IA’s effort to erase authors and publishers from the copyright system.

[1] The parties have consented to the filing of this brief. Amici Curiae and their counsel authored this brief. Neither a party, its counsel, nor any person other than Amici and their counsel contributed money that was intended to fund preparing or submitting this brief.

[2] See SPA-38 (“IA accepts that ReDigi forecloses any argument it might have under Section 109(a).”); Dkt. 60, Brief for Defendant-Appellant Internet Archive (hereinafter “IA Br.”) (appealing only the district court’s decision on fair use).

[3] See, e.g., ReDigi, 910 F.3d at 662; UMG Recordings, Inc. v. MP3.com, Inc., 92 F. Supp. 2d 349, 352 (S.D.N.Y. 2000); see also Disney Enters., Inc. v. VidAngel, Inc., 869 F.3d 848, 861–62 (9th Cir. 2017).

Dear Chair Khan, Commissioners Slaughter and Bedoya, and General Counsel Dasgupta,

The International Center for Law & Economics (ICLE) respectfully submits this letter in response to your June 29, 2023, NPRM regarding amendments to the premerger notification rules that implement the Hart-Scott-Rodino Antitrust Improvements Act (HSR Act) and to the Premerger Notification and Report Form and Instructions.

ICLE is a nonprofit, nonpartisan research center working to promote the use of law & economics methodologies to inform public policy debate. We have a long history of participation in regulatory proceedings relating to competition and antitrust law, including recent revisions to the merger guidelines[1] and the proposed revisions to the HSR premerger notification process.[2] We are consistently grateful for the opportunity to participate in proceedings such as these.

We write to express our concern about an important omission in the FTC’s proposed changes to the premerger notification form: its failure to address the requirements of the Regulatory Flexibility Act (RFA).[3] We appreciate your interest in this matter and the opportunity to share our concern with your offices.

This concern involves two legislative frameworks: the HSR premerger notification process and the requirements of the RFA. Under the HSR Act’s amendments to the Clayton Act, firms engaging in mergers above a statutorily defined minimum value[4]—including many that would involve smaller businesses to which the RFA applies—are required to provide information about a proposed merger to the FTC and the Department of Justice (DOJ) before the transaction can close. To bolster information gathering in merger enforcement, the FTC (with concurrence from the DOJ) proposed an extensive set of amendments to the filing process outlined by the HSR Act.[5] The proposed changes to the HSR process would dramatically expand the disclosure obligations for merging companies that meet the minimum valuation threshold.[6]

Under the RFA, federal agencies “shall prepare and make available for public comment an initial regulatory flexibility analysis. . . describ[ing] the impact of the proposed rule on small entities,”[7] except where “the head of the agency certifies that the rule will not, if promulgated, have a significant economic impact on a substantial number of small entities.”[8] If the agency believes the amendment will not have a substantial impact on small businesses, then it must provide “the factual basis” for its conclusion.[9]

The statement certifying that the proposed HSR changes in the NPRM won’t affect small businesses reads, in full:

Because of the size of the transactions necessary to invoke an HSR Filing, the premerger notification rules rarely, if ever, affect small entities. The 2000 amendments to the Act exempted all transactions valued at $50 million or less, with subsequent automatic adjustments to take account of changes in Gross National Product resulting in a current threshold of $111 million. Further, none of the proposed amendments expands the coverage of the premerger notification rules in a way that would affect small entities. Accordingly, the Commission certifies that these proposed amendments will not have a significant economic impact on a substantial number of small entities.[10]

Unfortunately, this is insufficient to satisfy the requirements of the RFA. Although the FTC stresses the $111 million HSR threshold to assert that small entities will not be affected, the Small Business Administration (SBA) “generally defines a small business as an independent business having fewer than 500 employees.”[11] The SBA also offers more detailed, industry-specific identification of small businesses.[12] Indeed, the NPRM cites to the SBA’s own standards, but these, too, do not align with the FTC’s “factual statement,” and it is not evident that the Commission sufficiently delved into those standards to understand their relevance for the size thresholds under the HSR Act.

Even a quick review of the SBA’s “Small Business Size Standards by NAICS Industry” table reveals that the SBA classifies the size of a firm based on either annual receipts or number of employees,[13] depending upon the characteristics of their industry.[14] Neither of these, it should go without saying, is the same as a “size-of-transaction” threshold under the HSR Act. Nor does the NPRM’s “factual basis” statement contain information sufficient to determine that there is any correlation between the SBA’s size thresholds and the size of a transaction (which typically represents something between the discounted present value of a firm’s expected returns under new ownership and current ownership over an indefinite time period).

Despite the FTC’s claim that the $111 million deal threshold will ensure that small businesses are not substantially affected, the agency’s own data from 2022 shows that nearly a quarter of all HSR filings covered transactions involving firms with sales of $50 million or less.[15] The same data shows that, out of the 3029 reported transactions in 2022, 513 involved firms with between $50 and $100 million in sales and 305 with between $100 and $150 million in sales. Here, again, the SBA’s metrics for identifying small businesses bear emphasis: where the SBA relies on dollar values instead of employee headcounts to define small businesses at all, it does so based on annual average receipts, not on the overall value of the firm.

This distinction underscores a point made in a letter filed by the App Association, a trade group representing small technology firms, that it is important not to conflate valuation with size.[16] A company, such as an innovative tech startup, can have a small number of employees but a high value based on projected sales, intellectual property, and forthcoming products. Indeed, the App Association notes that a number of its members are already subject to HSR disclosures and that that number can only increase under the proposed amendments.[17]

To provide further context regarding whether many of these deals involve small businesses, a 2013 CrunchBase dataset showed that the average successful American startup sold for $242.9 million.[18] Furthermore, the FTC’s 2022 HSR report highlights at least one challenged transaction involving a small business: Meta/Within.[19] Within, a virtual-reality startup with 58 employees, was acquired by Meta for $400 million.[20] Not only was there an HSR filing, but the FTC attempted to challenge the transaction—and lost in district court.[21]

Small businesses are clearly burdened by the HSR premerger notification requirements—and this burden would only increase under the proposed changes. By the FTC’s own estimate the new requirements would quadruple the hours required to prepare an HSR filing and raise costs by $350 million. By other, more realistic estimates, that increase in work hours would entail a cost of more than $1.6 billion[22]—or, indeed, considerably more.[23] There is no question that drastically increasing the cost of merger filings will make it much harder for small businesses to merge or be acquired, which is a primary form of success for small businesses.

Indeed, the NPRM’s proposed changes are, in part, specifically designed to affect small businesses. “Acquisitions of small companies can cause harm, including in sectors where competition occurs on a local level. . . . Thus, the Commission proposes several changes to expand the requirements for information related to prior acquisitions beyond what is currently required by Item 8.”[24] Furthermore, “given the difficulties in determining the value of small or nascent companies, the Commission believes it would be less burdensome for filers to report all acquisitions. . . .”[25]  Indeed, the FTC is aware of the potential burden on small businesses that such an approach would entail, but nevertheless aims to ensure that its proposed rules “still captur[e] acquisitions of entities worth less than $10 million.”[26]

And there is yet a further problem: These concerns take into account only the direct costs the NPRM would impose on small businesses. But, as the National Federation of Independent Business highlighted in 2023, several agencies have arguably failed to comply with the RFA by failing to consider indirect effects on small businesses.[27] Obviously, there is no such analysis provided here—and, indeed, as noted above, a clear intent of the NPRM is to affect the likelihood of small-business acquisition by reducing the incentive for firms to serially acquire small businesses. Doing so, of course, reduces funders’ incentives to invest in startups and small businesses and raises these companies’ cost of capital. Arguably that increase is itself a direct cost, but certainly its indirect effect is incredibly significant to the health of small businesses in the U.S.

The dramatic changes to the HSR premerger notification requirements proposed by the FTC have already created substantial uncertainty within the antitrust bar. Procedural defects such as failing to comply with the requirements of the RFA increase the likelihood that any rules adopted by the FTC will be challenged in court. This would increase the uncertainty (and thus the cost) surrounding the HSR process. This would be an unfortunate outcome. Fortunately, it is one that can be avoided if the FTC addresses these issues prior to finalizing its proposed rules.

[1] Geoffrey A. Manne, Dirk Auer, Brian Albrecht, Eric Fruits, Daniel J. Gilman, & Lazar Radic, Comments of the International Center for Law and Economics on the FTC & DOJ Draft Merger Guidelines, International Center for Law and Economics (Sept 18, 2023), https://laweconcenter.org/resources/comments-of-the-international-center-for-law-and-economics-on-the-ftc-doj-draft-merger-guidelines/.

[2] Brian Albrecht, Dirk Auer, Daniel J. Gilman, Gus Hurwitz, & Geoffrey A. Manne, Comments of the International Center for Law & Economics on Proposed Changes to the Premerger Notification Rules, International Center for Law and Economics (Sept 27,2023), https://laweconcenter.org/resources/comments-of-the-international-center-for-law-economics-on-proposed-changes-to-the-premerger-notification-rules/.

[3] 5 U.S.C. §§ 601-612 (2018).

[4] 15 U.S.C. § 18a(a)(2) (2018).

[5] NPRM, 88 FR 42178 (Jun. 29, 2023).

[6] See id. at 42208 (estimating the hours and expenses required to comply with the new rules). According to antitrust practitioners, however, the NPRM’s estimate likely substantially underestimates the true burden and cost of the proposed rules. See, e.g., Sean Heather, Antitrust Experts Reject FTC/DOJ Changes to Merger Process, Chamber of Commerce (Sept 19, 2023), https://www.uschamber.com/finance/antitrust/antitrust-experts-reject-ftc-doj-changes-to-merger-process.

[7] 5 U.S.C. § 603(a) (2018).

[8] 5 U.S.C. § 605(b) (2018).

[9] Id.

[10] NPRM, 88 FR 42178, 42208 (Jun. 29, 2023).

[11] Frequently Asked Questions, U.S. Small Bus. Admin. Off. of Advoc. (2023), https://advocacy.sba.gov/wp-content/uploads/2023/03/Frequently-Asked-Questions-About-Small-Business-March-2023-508c.pdf.

[12] See 13 CFR § 121.101, et seq. (1996)

[13] 13 CFR § 121.201 (2024).

[14] Indeed, the SBA’s standards entail a review of a wide range of such characteristics. See 13 CFR § 121.102 (1996) (“SBA considers economic characteristics comprising the structure of an industry, including degree of competition, average firm size, start-up costs and entry barriers, and distribution of firms by size. It also considers technological changes, competition from other industries, growth trends, historical activity within an industry, unique factors occurring in the industry which may distinguish small firms from other firms, and the objectives of its programs and the impact on those programs of different size standard levels.”).

[15] See Fed. Trade Comm’n and Dept of Just., Hart-Scott-Rodino Annual Report (2022), at Table IX, available at https://www.ftc.gov/system/files/ftc_gov/pdf/FY2022HSRReport.pdf.

[16] See Letter from Morgan Reed, President of App Association, to Lina Khan, Chair of Fed. Trade. Comm’n and Members of Congress (Feb 1, 2024), available at https://actonline.org/wp-content/uploads/App-Association-HSR-RFA-Ltr-1-Feb-2024-1.pdf.

[17] See id.

[18] See Mark Lennon, CrunchBase Reveals: The Average Successful Startup Raises $41M, Exits at $242.9M, TechCrunch (Dec 14, 2013), https://techcrunch.com/2013/12/14/crunchbase-reveals-the-average-successful-startup-raises-41m-exits-at-242-9m.

[19] See Fed. Trade Comm’n and Dept of Just., Hart-Scott-Rodino Annual Report (2022), available at https://www.ftc.gov/system/files/ftc_gov/pdf/FY2022HSRReport.pdf.

[20] See, e.g., Within (Virtual Reality) Overview, Pitchbook (last visited Feb. 29, 2024), https://pitchbook.com/profiles/company/117068-59#overview.

[21] In the Matter of Meta/Zuckerberg/Within, Fed. Trade Comm’n Docket No. 9411 (Aug. 11, 2022), https://www.ftc.gov/legal-library/browse/cases-proceedings/221-0040-metazuckerbergwithin-matter.

[22] See Albrecht, et al., supra note 2, at 7 (“The U.S. Chamber of Commerce conducted ‘a survey of 70 antitrust practitioners asking them questions about the proposed revisions to the HSR merger form and the new draft merger guides.’ Based on average answers from the survey respondents, the new rules would increase compliance costs by $1.66 billion, almost five times the FTC’s $350 million estimate.”).

[23] See id. (“For the current rules, the average survey response puts the cost of compliance at $79,569. Assuming there are 7,096 filings (as the FTC projects for FY 23), the total cost under the current rules would be $565 million. Under the new rules, the average survey response estimates the expected cost of compliance to be $313,828 per transaction, for a total cost of $2.23 billion.”) (emphasis added).

[24] NPRM, 88 FR 42178, 42203 (Jun. 29, 2023).

[25] Id. at 42204 (emphasis added).

[26] Id. (emphasis added).

[27] See Rob Smith, The Regulatory Flexibility Act: Turning a Paper Tiger Into a Legitimate Constraint on One-Size-Fits-All Agency Rulemaking, NFIB Small Business Legal Center (May 2, 2023), https://strgnfibcom.blob.core.windows.net/nfibcom/NFIB-RFA-White-paper.pdf (collecting examples).

ICLE Director of Law & Economics Programs Gus Hurwitz was a guest on The Cyberlaw Podcast, where he discussed big news for cord-cutting sports fans, Amazon’s ad-data deal with Reach, a novel Federal Trade Commission case brought against Blackbaud, the Federal Communications Commission’s ban on AI-generated voice cloning in robocalls, and South Korea’s pause on implementation of its anti-monopoly platform act. Audio of the full episode is embedded below.

Abstract

Standard accounts of why we have administrative agencies do little to account for those agencies’ ability to generate new information that can inform the regulatory process. Even expertise-based understandings of the administrative state limit the role of agencies to gathering information; and prevailing understandings of the administrative state view agencies as engaged in a policy-development exercise checked by theories of political accountability. This is unfortunate, because, for the same reasons that Congress turns to agencies to regulate in complex policy domains, agencies are typically in the best position to generate and make productive use of information that can inform the regulatory process and help Congress to accomplish its intended legislative goals.

This article offers a new account of how we can—and should—think about agencies’ use of information in the regulatory process: regulation as a discovery process. Drawing from economic understandings of how information is produced and used in both regulation and markets, it argues that using the regulatory process to generate information and ensuring that that information is both captured and productively used to improve regulations should be a priority for administrative law. In so doing, it contributes to a growing literature that argues for more experimentation in regulation and offers an account of the administrative state that is divergent from the interest group and presidential administrative models. Specific applications of these ideas are considered. These include how viewing regulation as a discovery process can resolve tensions in the Major Questions Doctrine and the use of an Executive Order to treat regulations as data-generating natural experiments.

As scholars, former judges, and former government officials who are experts in patent law, patent licensing, and innovation policy, we respectfully submit this comment in response to the Request for Information (RFI) by the National Institute of Standards and Technology (NIST) on the Draft Interagency Guidance Framework for Considering the Exercise of March-in Rights (Guidance Framework).[1] In the RFI, NIST states that it seeks “to ensure that [the Guidance Framework] is clear, and its application will both fulfill the purpose of march-in rights and uphold the policy and objectives of the Bayh-Dole Act.”[2] We believe that the Guidance Framework contradicts both the text and purpose of the Bayh-Dole Act, and thus it should be withdrawn by NIST.

For the first time since the enactment of the Bayh-Dole Act in 1980, NIST proposes a Guidance Framework for the four march-in powers in 35 U.S.C. § 203 that provides that “march-in is warranted” and thus an agency may issue licenses without authorization by the patent owner if “the price or other terms at which the product is currently offered to the public are not reasonable.”[3] The RFI expressly states that agencies that provided funding for subject inventions under the Bayh-Dole Act may “include consideration of factors that unreasonably limit availability of the invention to the public [as triggers of the march-in powers under § 203], including the reasonableness of the price and other terms at which the product is made available to end-users.”[4]

The Guidance Framework’s inclusion of “the reasonableness of the price [paid by] end-users” as a new criterion for any agency exercising the march-in powers in § 203 represents unprecedented and unauthorized regulatory authority. It lacks statutory authorization in the Bayh-Dole Act, as confirmed by its text, its purpose, and by other sources of statutory interpretation long relied on by courts and agencies, such as past interpretations of a statute by government officials. In fact, § 203 of the Bayh-Dole Act never mentions “price” as a criterion for the exercise of the four specified march-in powers, as contrasted with the RFI’s reference to “price” twenty-six (26) times.

Congress knows how to enact a price-control statute and to state clearly in a statute’s text that federal officials or agencies may consider “reasonable price” or even merely “price” as a condition for authorizing direct or indirect price controls on products produced and sold by private companies to consumers. One example is the Emergency Price Control Act of 1942,[5] among many others. The Bayh-Dole Act does not authorize this administrative power to control directly or indirectly prices, neither generally nor specifically in the four march-in conditions in § 203.

Other organizations and individuals with direct experience and knowledge in research and development in companies and universities, patent licensing under the Bayh-Dole Act, and in other related commercial activities in the U.S. innovation economy have submitted comments on these matters about which they have expertise. As legal experts, our comment explains why the Bayh-Dole Act does not authorize an agency to issue march-in licenses for the purpose of lowering prices on any product or service embodying a patent covered by this statute. First, it describes the evidence of the proven success of the patent system as a driver of innovation and economic growth. This is the necessary legal and policy framework for evaluating any proposed regulatory alterations to patent rights, especially unprecedented proposals like the Guidance Framework that would weaken or eliminate these patent rights. Second, it explains why the Guidance Framework lacks authorization in the Bayh-Dole Act according to its plain text, its statutory function, and its consistent implementation by agencies over several decades by bipartisan administrations. Third, it identifies how Senators Birch Bayh and Robert Dole expressly rejected claims by professors over two decades ago that the Bayh-Dole Act authorized agencies to use the march-in powers to control market prices of products and services. NIST should withdraw the proposed Guidance Framework.

The Success of the Patent System as a Driver of Economic Growth and Innovation

The patent system has been a key driver of the U.S. innovation economy for over 200 years, as economists, historians, and legal scholars have repeatedly demonstrated.[6] The patent system was central to the successes of the Industrial Revolution in the nineteenth century, the pharmaceutical and computer revolutions in the twentieth century, and the biotech and mobile telecommunications revolutions in the twenty-first century.[7] Patent systems that secure reliable and effective property rights to inventors consistently and strongly correlate with successful innovation economies.[8]

Dr. Zorina Khan, an award-winning economist, has demonstrated that reliable and effective property rights in innovation—patents—were a key factor in thriving markets for technology in the United States in the nineteenth century.[9] Other economists have also identified features of these robust nineteenth-century innovation markets—such as an increase in “venture capital” investment in patent owners, the rise of a secondary market in the sale of patents as assets, and the embrace of specialization via licensing business models—as indicators of value-maximizing economic activity made possible by reliable and effective patents.[10] This remains true today: a twenty-first-century startup with a patent more than doubles its chances of securing venture capital financing compared to a startup without a patent, and this patent-based startup has statistically-significant increased chances of success in the marketplace as well.[11]

These general economic insights and historical facts are especially evident in the biopharmaceutical sector. Historically, the U.S. has been a global leader in first securing innovations in new drugs, diagnostics, and other biotech innovations in healthcare.[12] As a result, the U.S. is a global leader in biomedical innovation. More than one-half of new drugs worldwide are invented in the U.S., improving the quality and duration of human life here and abroad.[13] For this reason, the U.S. patent system was identified as the “gold standard” in securing reliable and effective property rights in the fruits of innovative labors—patents.[14]

The real-world results of reliable and effective property rights—whether in real property or in patents—is extensive private investments, development of new products and services, and the creation and growth of new commercial markets. Just as in the high-tech sector and in the mobile revolution,[15] these same economic consequences are manifest in modern healthcare. The annual private investment in research and development (R&D) of new pharmaceutical and biotech innovations is approximately $129 billion (as of 2018).[16] This is almost triple the total amount of total public funding of $43 billion of R&D in healthcare innovations (as of 2018).[17] Medical diagnoses that once were either death sentences or led to a greatly diminished quality of life—cancer, hepatitis, and diabetes—are now treatable and manageable medical conditions within a relatively normal lifespan. This data is relevant in assessing the Guidance Framework because the Biden Administration has argued that it serves the purpose of lowering drug prices,[18] although the Guidance Framework does not state this nor does it limit the proposed “reasonable price” criterion to patented drugs and other inventions resulting from some upstream research funding in the life sciences by the federal government.

The evidence of the historical, economic, and empirical success of the U.S. patent system in driving innovation and economic growth is the baseline by which NIST should consider new regulatory proposals that ultimately weaken or restrict reliable and effective patents on new innovations throughout all sectors of the U.S. innovation economy. This includes the Guidance Framework, which includes an unprecedented power to issue nonexclusive licenses for the purpose of controlling prices on any patented product or service because a funding agency may deem it to be sold at “unreasonable prices.” The eight scenarios and examples in the Guidance Framework make clear that consideration of “reasonable price” as a condition for exercising the march-in power applies to every sector of the U.S. innovation economy, from manufacturing of highway signage to the 5G communication technologies implemented in connected cars.[19]

The evidentiary burden is on any official or agency proposing wide-ranging regulatory restrictions, additional costs, and additional legal uncertainties on patent owners. First, they must explain that proposed regulations are legally authorized. Second, they must explain, even if legally authorized, that there is reliable and robust data that supports this proposal as evidence-based policymaking. As will now be explained the Guidance Framework fails on both of these necessary conditions for an agency adopting new regulations, especially those that authorize unprecedented powers such as the Guidance Framework’s authorization of an agency to impose price controls under a “reasonable price” criterion for issuing nonexclusive licenses under § 203 of the Bayh-Dole Act. § 1498. These arguments are equally incorrect, as detailed below.

A Price-Control Power Contradicts the Text and Statutory Purpose of the Bayh-Dole Act

Congress enacted the Bayh-Dole Act in 1980 to provide an incentive for private parties to make the significant, risky investments in new product development, in creating manufacturing capabilities, and in setting up supply and distribution chains that bring new innovations to consumers. These are necessary investments in translating original discoveries into useful commercial products.[20] Before 1980, the government effectively claimed ownership in inventions resulting from government-funded research, offering nonexclusive licenses to anyone requesting one; this undermined the commercialization of these inventions given the absence of property rights that are the legal platform for contracts and other commercial activities.[21] The Bayh-Dole Act corrected this mistaken policy by establishing that innovators can obtain patents for inventions arising from some government-funded research and retain ownership in these patents, which facilitates licensing and other commercial activities in the marketplace.[22]

Section 203 in the Patent Act, as enacted in the Bayh-Dole Act, creates the limited exception to this core function of the Bayh-Dole Act by creating the “march in right.”[23] To ensure commercialization of inventions arising from research funded by government agencies, § 203 authorizes a federal agency that has funded research that resulted in a patented invention “to grant a nonexclusive, partially exclusive, or exclusive license” under four specified conditions.[24] A federal agency may grant these licenses “to a responsible applicant” without authorization from the patent owner in four delimited circumstances: (1) if “the contractor or assignee has not taken, or is not expected to take within a reasonable time, effective steps to achieve practical application of the subject invention in such field of use,” (2) “to alleviate health or safety needs which are not reasonably satisfied,” (3) “requirements for public use specified by Federal regulations . . . are not reasonably satisfied,” or (4) “a licensee of the exclusive right to use or sell any subject invention in the United States is in breach of its agreement.”[25]

The statutory text of § 203 does not support the unprecedented inclusion of “reasonable price” as a criterion for any agency in imposing price controls on patented products or services produced by private companies and sold to private consumers in the marketplace. The four march-in conditions, set forth in § 203(a) in the disjunctive, constitute the only authorizations in this exemption in the Bayh-Dole Act for a federal agency to exercise the march-in power. Notably, there is no mention of “reasonable price” in the four authorizing conditions for a federal agency to invoke the march-in power to issue licenses without approval from a patent owner.

Congress would have expressly enacted text conferring a price-control power in § 203 if it intended a “reasonable price” to trigger use of the march-in power under § 203. Congress has enacted numerous statutes that have authorized officials or agencies to impose price controls on transactions in the marketplace.[26] The Emergency Price Control Act of 1942 is one such example.[27] Similarly, rate-regulation statutes enacted by the states according to their police powers expressly authorize legislators or regulators to set “prices” or determine “rates.”[28] Contrary to these price-control or rate-regulation statutes, § 203 is devoid of any archetypical pricing terms, such as “price,” “prices charged by an assignee or licensee,” “market price,” or “reasonable price.” According to the “the ordinary meaning of the words used” in § 203 and § 201(f) in the Bayh-Dole Act, the march-in power does not authorize licenses for the purpose of imposing price controls.[29]

Moreover, there is no catch-all clause in § 203 authorizing the march-in power for anything not already covered by the four specific march-in conditions. This is significant for at least two reasons. First, Congress knows how to create broadly framed and expansive authorizations for agency action, if this is its purpose. For example, Congress has expressly created broadly-framed authorizations of general administrative powers in other statutes, such as the well-known language in the Federal Communications Act of 1934 authorizing the Federal Communications Commission to grant radio transmission licenses according to whether the “public convenience, interest, or necessity will be served thereby.”[30] Second, the canon of statutory construction of expressio unius est exclusio alterius establishes that, without a catch-all clause, the march-in power is delimited to only these four express exemptions from the longstanding rights of patent owners covered by the Bayh-Dole Act to freely assign or license their property in the marketplace.[31] In sum, Congress chose not to create an open-ended grant of authority in § 203 in listing only four specific march-in conditions that strictly specify the narrow scope and application of the march-in power exemption in the Bayh-Dole Act, which comports with the general function of the Bayh-Dole Act in promoting private commercialization of patented innovations in the marketplace.

The inclusion of “reasonable price” as a criterion in the Guidance Framework follows the work of activists and academics who have argued for over two decades that the first condition in the march-in provision that specifies the failure “to achieve practical application” of an invention as a trigger for the march-in power means that that prices can prevent this “practical application” with consumers.[32] As is typical of modern legislation, the Bayh-Dole Act has a lengthy definition of “practical application” in which these advocates for this price-control theory of § 203 have focused on a single phrase (“available to the public on reasonable terms”).[33] These activists and academics have spun an entire theory of unprecedented and vast regulatory power to control prices in the marketplace of patented products and services based on only two general phrases in two separate sections of the Bayh-Dole Act—“practical application” and “reasonable terms.”

This price-control theory of § 203 is wrong as a matter of law and statutory interpretation. First, their argument creates vast administrative powers based on an out-of-context, laser-like focus on phrases that have been isolated from lengthy and complex statutory provisions. This commits the classic interpretative error of wooden textualism.[34] For example, these activists and academics do not acknowledge that “terms” is often a distinct legal concept from “price,” as these distinct words have been used in many legal instruments. In fact, statutes often distinguish between “price” and “terms” by listing these two words separately.[35]

These advocates for the price-control theory of § 203 also do not acknowledge that the partial definition of “practical application” in § 203(a)(1) as “reasonable terms” in § 201(f) in the Bayh-Dole Act follows past usage of “practical application,” which was understood to refer to the “successful development and terms of the license, not with a product’s price.”[36] For example, President John F. Kennedy issued a statement on patent policy in 1963 in which he proposed mandating licensing of government-owned inventions in order to achieve “practical application” of an invention and to “guard against failure to practice the invention.”[37]

Second, in interpreting a specific statutory provision or a specific clause within a statutory provision, the advocates for the price-control theory of § 203 violate fundamental legal rules governing the interpretation and application of statutes. Courts always inquire into “the specific context in which that language is used, and the broader context of the statute as a whole.”[38] The Supreme Court has bluntly stated in far too many cases to cite or quote: “We do not . . . construe statutory phrases in isolation; we read statutes as a whole.”[39] “Courts have a ‘duty to construe statutes, not isolated provisions.’”[40]

Congress stated its express intent in the Bayh-Dole Act: “It is the policy and objective of the Congress to use the patent system to promote the utilization of inventions arising from federally supported research or development.”[41] The march-in power is an exemption from the function of the Bayh-Dole Act to stimulate universities and other researchers receiving federal research funds to obtain patents to utilize licenses in commercializing their inventions. In fact, this exemption was included in the Bayh-Dole Act precisely because it advanced this primary commercialization function of the statute: if a patented invention is not licensed or made available in the marketplace by its owner or licensees, then an agency is authorized to act to achieve this goal. Thus, § 203(a)(1)-(4) specifies four conditions in which the march-in power is justified, and these conditions identify situations in which inventions are not sold or commercialized in the marketplace.[42]

Lastly, the Guidance Framework’s lack of legal authorization in the Bayh-Dole Act is confirmed by Supreme Court precedent that agencies may not arrogate powers to themselves that are not specifically granted in statutes. An unprecedented power to impose price controls on all patented products or services produced and sold in the marketplace that were created from upstream research supported by some federal funding requires more than vague or generalized statutory terms like “effective steps to achieve practical application.” This is especially true given that Congress has consistently and repeatedly rejected bills that would impose compulsory licensing on U.S. patent owners, from the First Congress in 1790 up through the twentieth century.[43]

The Supreme Court has consistently instructed agencies that “Congress, we have held, does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions— it does not, one might say, hide elephants in mouseholes.”[44] The Supreme Court has rejected other agencies’ claims to regulatory authority under similarly vague and generalized terminology as the statutory phrase “practice application” in § 203, which has been the justification of the price-control power that the Guidance Framework implements. In these many other legal cases, the Supreme Court has stated bluntly that “‘Congress could not have intended to delegate’ such a sweeping and consequential authority ‘in so cryptic a fashion.’”[45] The Supreme Court again stated last year that it repeatedly “requires Congress to enact exceedingly clear language if it wishes to significantly alter . . . the power of the Government over private property.”[46] The Guidance Framework lacks a clear authorization in § 203 to justify its unprecedented inclusion of “reasonable price” as a criterion for authorizing the march-in power.

Agency Interpretations of § 203 Confirm It Does Not Authorize a Price-Control Power

The plain text of § 203 and its function within the Bayh-Dole Act as a whole explains why federal agencies—spanning bipartisan administrations over several decades—have repeatedly rejected numerous petitions to use the march-in power to impose price controls on drug patents. In 2016, the Congressional Research Service identified six petitions submitted to the NIH requesting it to exercise its march-in power solely for the purpose of lowering prices of patented drugs sold in the healthcare market.[47] The NIH denied all six petitions on the grounds that § 203, as confirmed by the NIH’s prior interpretation of this statutory provision, did not permit the march-in power to be used for the purpose of lowering drug prices.[48] By 2019, four more petitions had been filed with the NIH by policy organizations and activists, each requesting again that the NIH invoke the march-in power for the sole purpose of lowering drug prices.[49] As with the prior six petitions reaching back to the 1990s, the NIH rejected these petitions on the statutory ground that “the use of march-in to control drug prices was not within the scope and intent of its authority.”[50]

In 1997, for example, the NIH was petitioned to invoke the march-in power for the Isolex 300, a patented medical device used in organ transplant procedures.[51] The NIH rejected the petition for failing to meet the burden of proof that any of the four march-in conditions specified in § 203 had been triggered, authorizing the NIH to march in and license other companies to make and sell this medical device in the healthcare market. The NIH found that the Isolex 300 was being commercialized in the marketplace: the patent owner was actively licensing the patented device, seeking regulatory approval, and meeting research demands.[52] These facts precluded the triggering of the march-in power under the four authorizing conditions in § 203.

In rejecting this march-in petition, the NIH further explained why lowering prices on a medical device like the Isolex 300—imposing price controls on the healthcare market—was not justified by the plain text of § 203 and the function of the Bayh-Dole Act in promoting the commercialization of patented inventions. The NIH stated that, even if the petitioner proved that there would be greater accessibility and lower prices given additional licenses from the NIH invoking the march-in power, this rationale lacked authorization under § 203.[53] The NIH stated bluntly that the march-in power in § 203 did not exist for the purpose of “forced attempts to influence the marketplace.”[54] It acknowledged the contradiction between the Bayh-Dole Act’s primary function in promoting the commercialization of new innovations in the marketplace and adopting a march-in power for the purpose of imposing price controls, observing that “such actions may have far-reaching repercussions on many companies’ and investors’ future willingness to invest in federally funded medical technologies.”[55] This was not merely a freestanding policy assessment by the NIH of this petition; it derived this conclusion from the plain meaning of § 203 within the context of the Bayh-Dole Act and its commercialization function.

Another petition in 2004 again requested that the NIH invoke the march-in power in § 203 to license a patent specifically to lower the price for Norvir, a drug used to treat AIDS. Again, the NIH rejected the petition.[56] The NIH explained that “the extraordinary remedy of march-in is not an appropriate means of controlling prices,” and that “[t]he issue of drug pricing has global implications and, thus, is appropriately left for Congress to address legislatively.”[57] The NIH again rejected another march-in petition seeking to lower the price of Norvir in 2013, again stating that the imposition of price controls on drug patents was not a statutorily authorized march-in power in § 203 of the Bayh-Dole Act.[58] The NIH bluntly concluded: “As stated in previous march-in considerations the general issue of drug pricing is appropriately addressed through legislative and other remedies, not through the use of the NIH’s march-in authorities.”[59] The frustration by NIH officials with the serial petitions seeking to impose price controls on drug patents via the march-in provision in the Bayh-Dole Act is palpable.

Lastly, on March 21, 2023, the NIH rejected the latest petition (filed again) for this agency to invoke the march-in power solely to lower the price of Xtandi, a cancer drug covered by patent.[60] In its latest rejection of the price-control theory of the Bayh-Dole Act, the NIH reiterated that the “purpose of the Bayh-Dole Act is to promote commercialization and public availability of government-funded inventions.”[61] With this statutory framework and purpose in mind, the NIH expressly “found Xtandi to be widely available to the public on the market” and “[t]herefore, the patent owner, the University of California, does not fail the requirement of bringing Xtandi to practical application.”[62] The NIH further pointed out that this decision about Xtandi is consistent with its prior multiple rejections of march-in petitions also seeking to lower drug prices.[63] It also recognized that the administrative processes and delays, especially in light of Xtandi’s remaining patent term, led it to conclude that “NIH does not believe that use of the march-in authority would be an effective means of lowering the price of the drug.”[64]

The NIH’s multiple decisions over several decades in interpreting the scope of the march-in power granted to it under § 203 is significant evidence that the Bayh-Dole Act does not authorize NIST to include “reasonable price” as a criterion for agencies like the NIH to use the march-in power under § 203. The eleven or more decisions ranging from the 1990s through 2023 in which the NIH has consistently rejected march-in petitions requesting it impose price controls on drug patents under § 203 constitute “the well-reasoned views of the agencies implementing a statute [that] ‘constitute a body of experience and informed judgment to which courts and litigants may properly resort for guidance.’”[65]

Original Sponsors of the Bayh-Dole Act Stated Their Law Did Not Authorize Price Controls

The Guidance Framework’s inclusion of “reasonable price” as a criterion for applying the march-in power under § 203 is a statutory power that was allegedly discovered and argued for by two professors in a law journal article published more than two decades after the enactment of the Bayh-Dole Act.[66] When they later published an op-ed advancing their article’s argument, Senator Birch Bayh and Senator Robert Dole responded by expressly rejecting their theory that the Bayh-Dole Act authorized price controls as an essential tool of the march-in power in § 203.

Professors Peter Arno and Michael Davis published an op-ed in the Washington Post in 2002 restating their argument from their law journal article the year before that the Bayh-Dole Act mandates that patented inventions resulting from “federal funds will be made available to the public at a reasonable price.”[67] Professors Arno and Davis’ op-ed prompted a response from Senators Bayh and Dole, published as a letter to the editor in the Washington Post two weeks later:

Bayh-Dole did not intend that government set prices on resulting products. The law makes no reference to a reasonable price that should be dictated by the government. . . . The [Arno and Davis] article also mischaracterizes the rights retained by the government under Bayh-Dole. The ability of the government to revoke a license granted under the act is not contingent on the pricing of the resulting product or tied to the profitability of a company that has commercialized a product that results in part from government-funded research. The law instructs the government to revoke such licenses only when the private industry collaborator has not successfully commercialized the invention as a product.[68]

Although this letter does not have the same legal status as the canons of statutory interpretation and official interpretation and application of a statute, Senators Bayh and Dole make clear that the inclusion of “reasonable price” as a criterion authorizing the march-in power is unconnected to the text or purpose of their statute. The proposed Guidance Framework, ultimately born of the price-control theory spawned by Professors Arno and Davis, is an unprecedented assertion of agency power to control prices in private market transactions without a legal basis in the Bayh-Dole Act.

Conclusion

The Guidance Framework proposes the addition of “reasonable price” as an unprecedented criterion for exercising the march-in powers specified in § 203 of the Bayh-Dole Act. This is a legally unjustified and unauthorized arrogation of power by NIST. The Bay-Dole Act does not state in its plain text a congressional authorization for federal agencies to consider “reasonable price” as a criterion for imposing price controls on all Bayh-Dole patented products or services that are commercialized in the marketplace. In addition to lack of authorization in the plain text of § 203, the Guidance Framework’s inclusion of “reasonable price” as a march-in criterion contradicts the function of Bayh-Dole in promoting the commercialization of inventions by patent owners in the marketplace. The NIH has consistently and repeatedly confirmed this lack of statutory authorization in § 203 to impose price controls across bipartisan administrations over several decades in rejecting all march-in petitions seeking to impose price controls.

NIST states in its RFI, “[t]o date, no agency has exercised its right to march-in,” but it fails to acknowledge the numerous, repeated rejections by the NIH of march-in petitions seeking to impose price controls on drug patents. NIST should follow these repeated actions by the NIH, including in its most recent rejection of the Xtandi march-in petition less than a year ago, in applying the clear text and function of the Bayh-Dole Act. Thus, NIST should withdraw the proposed Guidance Framework and permit the Bayh-Dole Act to function according to its intended function in promoting the commercialization of innumerable innovations in the marketplace.

[1] See National Institute of Standards and Technology, Request for Information Regarding the Draft Interagency Guidance Framework for Considering the Exercise of March-In Rights, 88 Fed. Reg. 85593 (Dec. 7, 2023).

[2] 88 Fed. Reg. 85593.

[3] Id. at 85598.

[4] Id. (emphasis added).

[5] See Pub. L. No. 77-421, 56 Stat. 23 (1942); see also Economic Stabilization Act of 1970, Pub. L. No. 91-379, § 202, 84 Stat. 799, 799-800 (“The President is authorized to issue such orders and regulations as he may deem appropriate to stabilize prices, rents, wages, and salaries at levels not less than those prevailing on May 25, 1970.”); Housing and Rent Act of 1947, Pub. L. No. 129, 61 Stat. 193, 198 (imposing rent controls on existing structures set at levels permitted to be charged under the Economic Price Control Act of 1942).

[6] See, e.g., ROBERT P. MERGES, AMERICAN PATENT LAW: A BUSINESS AND ECONOMIC HISTORY (2023); JONATHAN M. BARNETT, INNOVATORS, FIRMS, AND MARKETS: THE ORGANIZATIONAL LOGIC OF INTELLECTUAL PROPERTY (2021); DANIEL SPULBER, THE CASE FOR PATENTS (2021); B. ZORINA KHAN, INVENTING IDEAS: PATENTS, PRIZES, AND THE KNOWLEDGE ECONOMY (2020); Stephen Haber, Innovation, Not Manna from Heaven (Hoover Institution, Sep. 15, 2020); B. Zorina Khan, Trolls and Other Patent Inventions: Economic History and the Patent Controversy in the Twenty-First Century, 21 GEO. MASON L. REV. 825, 837-39 (2014); Naomi R. Lamoreaux, Kenneth L. Sokoloff & Dhanoos Sutthiphisal, Patent Alchemy: The Market for Technology in US History, 87 BUS. HIST. REV. 3 (Spring 2013); RONALD A. CASS & KEITH N. HYLTON, LAWS OF CREATION: PROPERTY RIGHTS IN THE WORLD OF IDEAS (2013).

[7] See generally MERGES, supra note 6; BARNETT, supra note 6; KHAN, supra note 6.

[8] See, e.g., Stephen Haber, Patents and the Wealth of Nations, 23 GEO. MASON L. REV. 811 (2016); Jonathan M. Barnett, Patent Tigers: The New Geography of Global Innovation, 2 CRITERION J. INNOVATION 429 (2017).

[9] See B. ZORINA KHAN, THE DEMOCRATIZATION OF INVENTION: PATENTS AND COPYRIGHTS IN AMERICAN ECONOMIC DEVELOPMENT, 1790–1920, at 9-10 (2005) (“[P]atents and . . . intellectual property rights facilitated market exchange, a process that assigned value, helped to mobilize capital, and improved the allocation of resources. . . . Extensive markets in patent rights allowed inventors to extract returns from their activities through licensing and assigning or selling their rights.”).

[10] See, e.g., Naomi R. Lamoreaux, Kenneth L. Sokoloff & Dhanoos Sutthiphisal, Patent Alchemy: The Market for Technology in US History, 87 BUS. HIST. REV. 3, 4–5 (2013).

[11] See Joan Farre-Mensa, et al., What Is a Patent Worth? Evidence from the U.S. Patent “Lottery,” 75 J. Finance 639 (2019), https://doi.org/10.1111/jofi.12867.

[12] See Kevin Madigan & Adam Mossoff, Turning Gold to Lead: How Patent Eligibility Doctrine Is Undermining U.S. Leadership in Innovation, 24 Geo. Mason L. Rev. 939, 942-44 (2017).

[13] See Ross C. DeVol, Armen Bedroussian & Benjamin Yeo, The Global Biomedical Industry: Preserving U.S. Leadership 5 (Sep. 2011), http://www.ncnano.org/CAMIExecSum.pdf.

[14] Madigan & Mossoff, supra note 12, at 940-41.

[15] See Letter from Alden Abbott, Kristina M.L. Acri, et al. to Assistant Attorney General Jonathan Kanter, Nov. 30, 2022, https://s3.amazonaws.com/media.hudson.org/Letter+to+AAG+Kanter+re+SEPs+and+Patent+Pools+10.30.22.pdf, at 1-2 (detailing economic evidence); see also Alexander Galetovic, Stephen H. Haber & Ross Levine, An Empirical Examination of Patent Holdup, 11 J. COMP. L. & ECON. 549, 564-69 (2015), https://papers.ssrn.com/abstract=2588169 (finding quality-adjusted prices for devices and other products in the patent-intensive telecommunications market to have fallen at a faster rate as compared to other sectors of the innovation economy).

[16] See U.S. Investments in Medical and Health Research and Development 2013–2018, at 7 (Research America, 2019), https://www.researchamerica.org/wp-content/uploads/2022/09/InvestmentReport2019_Fnl.pdf (estimating total private investment in biopharmaceutical R&D in 2018 is estimated to be $129 billion). For each drug approved by the FDA for use by patients, there is on average $2.6 billion in R&D expenditures incurred over 10–15 years. See Joseph A. DiMasi, Henry G. Grabowski, & Ronald W. Hansen, Innovation in the Pharmaceutical Industry: New Estimates of R&D Costs, 47 J. Health Econ. 20 (2016).

[17] See U.S. Investments in Medical and Health Research and Development 2013–2018, supra note 17, at 8.

[18] See FACT SHEET: Biden-?Harris Administration Announces New Actions to Lower Health Care and Prescription Drug Costs by Promoting Competition (Dec. 7, 2023), https://www.whitehouse.gov/briefing-room/statements-releases/2023/12/07/fact-sheet-biden-harris-administration-announces-new-actions-to-lower-health-care-and-prescription-drug-costs-by-promoting-competition/ (“Today, the Biden-Harris Administration is announcing new actions to promote competition in health care and support lowering prescription drug costs for American families, including the release of a proposed framework for agencies on the exercise of march-in rights on taxpayer-funded drugs and other inventions, which specifies that price can be a factor in considering whether a drug is accessible to the public.”).

[19] See 88 Fed. Reg. 85601-85605 (detailing the eight scenarios in which the march-in power may be used by an agency).

[20] See generally BARNETT, supra note 6.

[21] See, e.g., S. Rep. No. 480, 96th Cong., 1st Sess., at 2 (1979) (explaining that the government’s policy of owning patents on inventions arising from government-funded research and offering nonexclusive licenses “has proven to be an ineffective policy” and that “the private sector simply needs more protection for the time and effort needed to develop and commercialize new products than is afforded by a nonexclusive license”).

[22] See id., at 28 (“It is essentially a waste of public money to have good inventions gathering dust on agencies’ shelves because of unattractiveness of nonexclusive licenses.”).

[23] See 35 U.S.C. § 203 (2011).

[24] § 203(a).

[25] § 203(a)(1)-(4).

[26] See, e.g., Economic Stabilization Act of 1970, Pub. L. No. 91-379, § 202, 84 Stat. 799, 799-800 (“The President is authorized to issue such orders and regulations as he may deem appropriate to stabilize prices, rents, wages, and salaries at levels not less than those prevailing on May 25, 1970.”); Housing and Rent Act of 1947, Pub. L. No. 129, 61 Stat. 193, 198 (imposing rent controls on existing structures set at levels permitted to be charged under the Economic Price Control Act of 1942).

[27] See Pub. L. No. 77-421, 56 Stat. 23 (1942).

[28] See, e.g., Nebbia v. People of New York, 291 U.S. 502, 515 (1934) (“The Legislature of New York established by chapter 158 of the Laws of 1933, a Milk Control Board with power, among other things to ‘fix minimum and maximum … retail prices to be charged by … stores to consumers for consumption off the premises where sold.’”); Stone v. Farmers’ Loan & Trust Co., 116 U.S. 307, 308 (1886) (reviewing “the statute of Mississippi passed March 11, 1884, entitled ‘An act to provide for the regulation of freight and passenger rates on railroads in this state, and to create a commission to supervise the same, and for other purposes’”).

[29] INS v. Phinpathya, 464 U.S. 183, 189 (1984) (stating that “in all cases involving statutory construction, our starting point must be the language employed by Congress, . . . and we assume that the legislative purpose is expressed by the ordinary meaning of the words used”) (quotations and citations omitted).

[30] 47 U.S.C. § 307(a) (“The Commission, if public convenience, interest, or necessity will be served thereby, subject to the limitations of this Act, shall grant to any applicant therefor a station license provided for by this Act.”).

[31] See Tennessee Valley Authority v. Hill, 437 U.S. 153, 188 (1976) (“In passing the Endangered Species Act of 1973, Congress was also aware of certain instances in which exceptions to the statute’s broad sweep would be necessary. Thus, § 10, 16 U.S.C. § 1539 (1976 ed.), creates a number of limited ‘hardship exemptions,’ . . . . meaning that under the maxim expressio unius est exclusio alterius, we must presume that these were the only ‘hardship cases’ Congress intended to exempt.”); see also 73 Am. Jur. 2d Statutes § 129 (2002) (describing the statutory canon of interpretation, expressio unius est exclusio alterius).

[32] See, e.g., Letter from Amy Kapczynski, Aaron S. Kesselheim, et al. to Senator Elizabeth Warren, at 6-7 (Apr. 20, 2022), https://tinyurl.com/yt62wt4t; Fran Quigley & Jennifer Penman, Better Late than Never: How the U.S. Government Can and Should Use Bayh-Dole March-In Rights to Respond to the Medicines Access Crisis, 54 WILLAMETTE L. REV. 171 (2017); Peter S. Arno & Michael H Davis, Why Don’t We Enforce Existing Drug Price Controls? The Unrecognized and Unenforced Reasonable Pricing Requirements Imposed upon Patents Deriving in Whole or in Part from Federally Funded Research, 75 TULANE L. REV. 631 (2001).

[33] See 35 U.S.C. § 201(f) (defining “practical application” to mean “to manufacture in the case of a composition or product, to practice in the case of a process or method, or to operate in the case of a machine or system; and, in each case, under such conditions as to establish that the invention is being utilized and that its benefits are to the extent permitted by law or Government regulations available to the public on reasonable terms”).

[34] See Sackett v. Environmental Protection Agency, 143 S. Ct. 1322, 1340 (2023) (“construing statutory language is not merely an exercise in ascertaining ‘the outer limits of a word’s definitional possibilities’”) (quoting FCC v. AT&T, 562 U.S. 397, 407 (2011)); cf. Antonin Scalia, Common-Law Courts in a Civil Law System: The Role of the United States Federal Courts in Interpreting the Constitution and Law, in A MATTER OF INTERPRETATION: FEDERAL COURTS AND THE LAW 23-24 (Amy Gutmann, ed., 1997) (critiquing out-of-context linguistic construction of statutory terms because a “good textualist is not a literalist”).

[35] See, e.g., 47 U.S.C. § 335(b)(3) (“A provider of direct broadcast satellite service shall meet the requirements of this subsection by making channel capacity available to national educational programming suppliers, upon reasonable prices, terms, and conditions, as determined by the Commission . . . .”) (emphasis added); 42 U.S.C. § 2375 (“The charges and terms for the transfer of any utility may be established by advertising and competitive bid, or by negotiated sale or other transfer at such prices, terms, and conditions as the Commission shall determine to be fair and equitable.”) (emphases added); 10 U.S.C. § 3372(a)(1) (“A contracting officer of the Department of Defense may not enter into an undefinitized contractual action unless the contractual action provides for agreement upon contractual terms, specifications, and price . . . .”) (emphasis added); 43 U.S.C. § 375c (“The Secretary is authorized to sell such land to resident farm owners or resident entrymen, on the project upon which such land is located, at prices not less than that fixed by independent appraisal approved by the Secretary, and upon such terms and at private sale or at public auction as he may prescribe . . . .”) (emphases added); 2 U.S.C. § 4103 (“[I]n any contract which is entered into by any person and either the Administrator of General Services or a contracting officer of any executive agency and under which such person agrees to sell or lease to the Federal Government (or any one or more entities thereof) any unit of property, supplies, or services at a specified price or under specified terms and conditions (or both), such person may sell or lease to the Congress the same type of such property, supplies, or services at a unit price or under terms and conditions (or both) . . . .”) (emphases added).

[36] Joseph Allen, New Study Shows Bayh-Dole is Working as Intended—and the Critics Howl, IPWATCHDOG (March 12, 2019), https://www.ipwatchdog.com/2019/03/12/new-study-shows-bayh-dole-working-intended/id=107225/.

[37] Government Patent Policy, Memorandum of Oct. 10, 1963, Fed. Reg. 10943 (Oct. 12, 1963).

[38] Robinson v. Shell Oil Co., 519 U.S. 337, 340 (1997).

[39] Samantar v. Yousuf, 560 U.S. 305, 319 (2010) (quoting United States v. Morton, 467 U.S. 822, 828, (1984)).

[40] Graham Cty. Soil & Water Conservation Dist. v. U.S. ex rel. Wilson, 559 U.S. 280, 290 (2010) (quoting Gustafson v. Alloyd Co., 513 U.S. 561, 568 (1995)); see also Gonzales v. Oregon, 546 U.S. 243, 273 (2006) (stating that “statutes ‘should not be read as a series of unrelated and isolated provisions.’”) (quoting Gustafson v. Alloyd Co., 513 U.S. 561, 570, (1995)); Food & Drug Admin. v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 133 (2000) (“It is a ‘fundamental canon of statutory construction that the words of a statute must be read in their context and with a view to their place in the overall statutory scheme.’”) (quoting Davis v. Michigan Dept. of Treasury, 489 U.S. 803, 809 (1989)); Louisville & N.R. Co. v. Gaines, 3 F. 266, 276 (C.C.M.D. Tenn. 1880) (“Where the language [of a statute] is clear and explicit the court is bound . . . . It must be construed as a whole. The office of a good expositor, says My Lord Coke, ‘is to make construction on all its parts together.’”).

[41] 35 U.S.C. § 200.

[42] See supra notes 23-31, and accompanying text.

[43] See, e.g., Bruce W. Bugbee, Genesis of American Patent and Copyright Law 143-44 (1967) (discussing the rejection of a Senate proposal for a compulsory licensing requirement in the bill that eventually became the Patent Act of 1790); Kali Murray, Constitutional Patent Law: Principles and Institutions, 93 Nebraska Law Review 901, 935-37 (2015) (discussing 1912 bill that imposed compulsory licensing on patent owners who are not manufacturing a patented invention, which received twenty-seven days of hearings, but was not enacted into law).

[44] Whitman v. Am. Trucking Associations, 531 U.S. 457, 468 (2001).

[45] See West Virginia v. Environmental Protection Agency, 142 S. Ct. 2587, 2608 (2022) (quoting Food & Drug Admin. v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 159 (2000)). See also MCI Telecommunications Corp. v. American Tel. & Tel. Co., 512 U.S. 218, 231 (1994) (“It is highly unlikely that Congress would leave the determination of whether an industry will be entirely, or even substantially, rate-regulated to agency discretion—and even more unlikely that it would achieve that through such a subtle device as permission to ‘modify’ rate-filing requirements.”).

[46] Sackett, 143 S. Ct. at 1341 (quoting United States Forest Service v. Cowpasture River Preservation Ass’n, 140 S. Ct. 1837, 1849-50 (2020)).

[47] See John R. Thomas, March-In Rights Under the Bayh-Dole Act 8-10 (Congressional Research Service, Aug. 22, 2016).

[48] Id.

[49] See Return on Investment Initiative for Unleashing American Innovation 29 (NIST Special Publication 1234, April 2019) (identifying 10 petitions to break patents through the march-in power in § 203 solely for the purpose of imposing price controls on drug patents).

[50] Id.

[51] See, e.g., NIH Office of the Director, Determination in the Case of Petition of CellPro, Inc. (Aug. 1, 1997), https://www.ott.nih.gov/sites/default/files/documents/policy/cellpro-marchin.pdf (rejecting petition in part to invoke march-in power given argument that company was too slow in bringing a medical device to market).

[52] Id.

[53] Id.

[54] Id. at 7.

[55] Id. at 7.

[56] See NIH Office of the Director, In the Case of Norvir Manufactured by Abbott Laboratories, Inc. (July 29, 2004), http://www.ott.nih.gov/sites/default/files/documents/policy/March-In-Norvir.pdf.

[57] Dr. Elias A. Zerhouni, Nat’l Institute of Health, Determination in the Case of Norvir I, at 5-6 (July 2, 2004).

[58] NIH Office of the Director, In the Case of Norvir Manufactured by AbbVie (Nov. 1, 2013), https://www.ott.nih.gov/sites/default/files/documents/policy/March-In-Norvir2013.pdf.

[59] Id.

[60] See Letter from Lawrence A. Tabak, Performing the Duties of the NIH Director, to Robert Sachs and Clare Love (Mar. 23, 2023), https://www.keionline.org/wp-content/uploads/NIH-rejection-Xtandi-marchin-12march2023.pdf (rejecting petition to impose price controls on Xtandi).

[61] Id. at 2.

[62] Id.

[63] Id.

[64] Id.

[65] See United States v. Mead Corp., 533 U.S. 218, 227 (2001) (quoting Bragdon v. Abbott, 524 U.S. 624, 642 (1998) (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)))

[66] See Arno & Davis, supra note 32.

[67] See Peter Arno & Michael Davis, Paying Twice for the Same Drugs, Washington Post (March 27, 2002), https://www.washingtonpost.com/archive/opinions/2002/03/27/paying-twice-for-the-same-drugs/c031aa41-caaf-450d-a95f-c072f6998931/ (emphasis added).

[68] Birch Bayh and Robert Dole, Our Law Helps Patients Get New Drugs Sooner, Wash. Post (Apr. 11, 2002), https://www.washingtonpost.com/archive/opinions/2002/04/11/our-law-helps-patients-get-new-drugs-sooner/d814d22a-6e63-4f06-8da3-d9698552fa24/.

ICLE Director of Law & Economics Programs Gus Hurwitz joined an online panel hosted by the Technology Policy Institute on where telecom industry observers can expect to see movement in spectrum reauthorization, mid-band clearing, and results from newly deployed shared bands. Video of the full event is embedded below.

We thank the Federal Communications Commission for this opportunity to comment on the rules proposed in the above-captioned Notice of Proposed Rulemaking.[1] We are both legal academics with long-standing interest in the topics addressed by these proposed rules. Over the past 15 years, we have each published numerous articles, books, and other scholarly works on network neutrality and related topics, submitted comments in relevant Commission proceedings and briefs in related judicial proceedings, and been active participants in academic and public discussion of these topics.[2]

The views presented below are ours alone and should not be attributed to our employer or to the Center for Technology, Innovation, and Competition. Neither of us has received any compensation for these comments, nor has either of us been retained by any party with a financial interest in these proceedings.

The essence of the rules – their purpose and flaws – are effectively captured in the first paragraphs of the NPRM:

1. Today we propose to reestablish the Federal Communications Commission’s (Commission) authority over broadband Internet access service by classifying it as a telecommunications service under Title II of the Communications Act of 1934, as amended (Act). While Internet access has long been important to daily life, the COVID-19 pandemic and the rapid shift of work, education, and health care online demonstrated how essential broadband Internet connections are for consumers’ participation in our society and economy. Congress responded by investing tens of billions of dollars into building out broadband Internet networks and making access more affordable and equitable, culminating in the generational investment of $65 billion in the Infrastructure Investment and Jobs Act.
2. But even as our society has reconfigured itself to do so much online, our institutions have fallen behind. There is currently no expert agency ensuring that the Internet is fast, open, and fair. Since the birth of the modern Internet in the 1990s, the Commission had played that role, but the Commission abdicated that responsibility in 2018, just as the Internet was becoming more vital than ever.

While the NPRM would establish various specific rules (e.g., the proposed conduct rules and transparency rule) and puts forth various justifications for them, the primary and most significant goal of the proposal is to classify Broadband Internet Access Service (BIAS) as a Title II telecommunications service. That is literally the first sentence of the NPRM.

This purpose is also the most objectionable and least necessary aspect of the proposed rules. Also as noted in the first paragraph, and discussed in the NPRM, the COVID-19 pandemic made more clear than ever the importance of the Internet to modern life. But unrecognized by the NPRM, this period also demonstrated the extent to which the pervasive regulation contemplated by Title II is unnecessary for the Internet to satisfy this important role. While stressed, and while benefitting from support from programs like the Emergency Broadband Benefit,[3] American Rescue Plan Act,[4] and Affordable Connectivity Program,[5] the Internet proved its importance throughout the pandemic largely by living up to the tasks to which it was unexpectedly put. A lack of Open Internet regulations did not prevent it from doing so.

The first paragraph also recognizes Congress’s investment of $65 billion to support broadband deployment in response to the pandemic.[6] In so doing, Congress tasked an agency other than the Commission, the National Telecommunications and Information Agency (NTIA), with primary responsibility for overseeing use of this funding through the Broadband Equity, Access, and Deployment (BEAD) program.[7] And Congress did not include any network neutrality or open Internet provisions as requirements for allocation of this funding. This calls into substantial question whether Congress views the Commission as the regulatory agency with regulatory authority as relates to the Internet and, especially, whether Congress views rules such as proposed in the NPRM as necessary.

Even to the extent that it is true that the Commission is a relevant expert agency, the assertions in the second paragraph of the NPRM are false. The Federal Trade Commission (FTC) has expertise in antitrust and consumer protection. To the extent that Internet Service Providers (ISP) commit to providing fast, open, and fair service to their users—promises routinely made—there is, in fact, an agency with relevant expertise to ensure those commitments are kept. And to the extent that an ISP does not make such commitments, the First Amendment limits the FCC’s ability to impose them upon the ISP through the proposed rules. And to the extent that the FCC has played that role “since the birth of the modern Internet in the 1990s,” Title II classification has only been the basis for that role for the brief period between adoption of the 2015 Open Internet Order and its recission in the 2018 Restoring Internet Freedom Order.

The Commission is needlessly steering into political controversy through Title II reclassification. The D.C. Circuit’s 2014 opinion in Verizon makes clear that implementation of the substantive rules proposed in the NPRM could be accomplished using the Commission’s Title I authority.[8] There would likely be widespread support for (or at least acceptance of) rules adopted on this basis, both political and from industry.

This concern is heightened by the current judicial landscape. The Supreme Court’s evolving Major Questions Doctrine jurisprudence casts significant doubt on the Commission’s ability to assert long-disclaimed pervasive regulatory authority over Internet services—services that the Commission has repeatedly referred to as of substantial economic and political significance.[9] This is heightened by the changing technology of Internet access, which calls into doubt the ongoing relevance of the analysis in Brand X[10], the most important precedent supporting Title II classification of BIAS.[11] And the Supreme Court is hearing this term cases that bear directly on First Amendment central to the proposed rules.[12] The proposed rules significantly implicate speech regulation and are therefore highly sensitive to changing First Amendment jurisprudence. Proposing rules prior to resolution of these ongoing cases creates a needles risk of uncertainty over the coming years—a risk that directly contradicts the importance that the Commission ascribes to the proposed rules.

The remainder of these comments further develop these concerns in four sections. Part I discusses how technological developments over the past twenty years raise doubt about the continued application of the conclusions in Brand X. Part II looks at the current state of the broadband Internet services industry, focusing on the BEAD Program. Part III looks at First Amendment considerations relating to the NPRM. And Part IV considers the Supreme Court’s recent Major Questions Doctrine jurisprudence.

I. Changes in BIAS Architecture Cast Doubt on Continued Relevance of Brand X

The most important precedent governing classification of BIAS as a Title I information service or Title II telecommunication service is the Supreme Court’s opinion in Brand X. The question in Brand X was, in a sense, the opposite of that presented by the NPRM.

Both involve the classification of broadband Internet service as either an “information service” (regulated under Title I of the Communications Act) or a “telecommunications service” (regulated under Title II of the Communications Act). These terms are defined in 47 U.S.C. 153:

(20) INFORMATION SERVICE.—The term “information service” means the offering of a capability for generating, acquiring, storing, transforming, processing, retrieving, utilizing, or making available information via telecommunications, and includes electronic publishing, but does not include any use of any such capability for the management, control, or operation of a telecommunications system or the management of a telecommunications service.

(43) TELECOMMUNICATIONS.—The term “telecommunications” means the transmission, between or among points specified by the user, of information of the user’s choosing, without change in the form or content of the information as sent and received.

(46) TELECOMMUNICATIONS SERVICE.—The term “telecommunications service” means the offering of telecommunications for a fee directly to the public, or to such classes of users as to be effectively available directly to the public, regardless of the facilities used.

In Brand X, the Court recognized that some portion of the Internet service offered by a cable ISP is a telecommunications service and that some other portion is an information service.[13] The question the Court faced was whether those were so tightly integrated that the entire offering could be classified as a single information service offering, as opposed to separate and distinct telecommunications and information service offerings.[14]

In the proposed rules, the Commission proposes to do the exact opposite: they would classify an offering that includes some amount of information services alongside a telecommunications service as an integrated telecommunications service. While it seems natural to characterize these as obverse sides of a coin, the statutory reality does not allow this approach. The definition of an “information service” is the offering of various capabilities “via telecommunications”—information services are, in other words, necessarily integrated with a telecommunications service. The definition of “telecommunications service,” on the other hand, includes exclusively the provision of telecommunications. Any blending of these services can only be an information service.

If the Commission is to classify BIAS as a Title II telecommunications service, such classification must apply solely to the portion of an ISP’s service that is not an information service. If, in other words, Domain Name System (DNS) services are information services, offering consumers a service that includes DNS access must necessarily fall outside of the scope of BIAS offerings classified as Title II services. The same applies for any other services offered by an ISP that are necessarily information services.

As noted in the NPRM, a similar argument was rejected by the D.C. Circuit Court of Appeals in its affirmation of the 2015 Open Internet Order’s classification of BIAS as a Title II service.[15] There, a two-judge majority accepted the Commission’s characterization of DNS service as relating to the “management, control, or operation of a telecommunications system or the management of a telecommunications service,” features which remove the service from the statutory definition of an information service.[16] Dissenting, Judge Williams would have rejected that argument as arbitrary and capricious.[17]

Even accepting the Commission’s argument, this approach turns the question into a technical one. Judge Williams notes that with this approach “the Commission set for itself a highly technical task of classification”[18]—an echo of the Supreme Court’s statement in Brand X that the question of whether transmission services and DNS service are functionally integrated “turns not on the language of the Act, but on the factual particulars of how Internet technology works and how it is provided.”[19]

The question of whether DNS service is a management, control, or operation function was briefly discussed in Brand X, as well—though it was not properly before or decided by the Court. Rather, it was raised by Justice Scalia in his dissent, to which the majority responded in its footnote 3. In his dissent, Justice Scalia characterizes DNS service as “scarcely more than routing information, which is expressly excluded from the definition of ‘information service.’”[20] In its footnote, the majority notes that “routing information” is not, in fact, among the functions excluded from the definition of an information service.[21] Of course, ISPs do offer routing services. They do not merely transmit information of the user’s choosing between or among points specified by the user without change in the form or content sent or received—they very often select for the user which endpoints to which the user’s information will be sent or from which it will be retrieved and the routes by which that information will be sent and received. And this is very often accomplished using DNS configurations.[22]

This is significantly more true today than it was at the time of the DC Circuit’s review of the 2015 Open Internet Order, and even more so true than it was at the time of the Brand X case. The highly technical nature of the Internet, including the features provided by some, but not all, ISPs, has changed dramatically since the 2015 Open Internet Order—and it continues to change. Some ISPs are working to deploy technologies like the Internet Engineering Task Force’s proposed Non-Queue-Building Per-Hop Behavior standard to improve DNS performance.[23] Some ISPs are working to support development of new active queue management technologies, such as the Internet Engineering Task Force’s proposed Low Latency, Low Loss, Scalable Throughput standard, which, again, have the potential to dramatically improve performance of Internet services.[24] Larger ISPs can implement dynamic routing across their networks to improve performance.[25] These are all service that some, but not all, ISPs are likely to implement to offer than enhance users’ ability to acquire, retrieve, use, or make available information—and in their improved forms none is necessary to the management or operation of the underlying telecommunications service.

Similarly, the necessity of DNS services provided by ISPs has only decreased in recent years. Today there are several free, open DNS services, including services provided by Google and Cloudflare. Statistics suggest that at least 20% of Internet users use these services instead of DNS services provided by their ISP.[26] There are parts of the world in which such services account for more than 50% of the market, demonstrating the extent to which such services are not necessary to the management or operation of the underlying telecommunications system.[27] Private VPN services are widely advertised and obviate the need for an ISP’s DNS services. Web browsers such as Chrome and Firefox include support for DNS over HTTPS, including their own DNS server settings—indeed, by default Firefox bypasses the ISPs’ DNS servers entirely.[28]

Together, these demonstrate the extent to which even a mundane-seeming service like DNS is more than a mere network management feature. While Justice Scalia’s argument in Brand X that DNS is not an information service might have been reasonable at the time, with each passing year that argument becomes weaker. Today, DNS is a standalone service, undergoing active research and development, offered in a competitive marketplace. Consumers can, and do, bypass their ISPs’ DNS services; standalone products like web browsers and VPN services can, and do, bypass their users’ ISPs. And ISPs invest in improving the quality of their own DNS services not because it improves their management of the service—marginal increases in DNS performance offer zero benefit to the management, control, or operation of a telecommunications system—but because it enhances the users’ ability to generate, acquire, store, transform, process, retrieve, utilize, or make available information via telecommunications.

II. Ongoing Development of the Broadband Industry, Notably Including Congressional Broadband Programs, Cast Doubt on the Proposed Rules

The NPRM expresses concern that ISPs “have the incentive and ability to engage in practices that pose a threat to Internet openness” and seeks comment on the state of competition in the market.[29] Contrary to the NPRM’s concerns, the state of competition in the BIAS market is robust and continues to increase. While the NPRM cites to 2021 data that that approximately 36 percent of households do not have access to two or more providers offering 100/20 Mbps wireline Internet service, that same data shows that more than 86 percent of households do have competitive options at the 25/3 Mbps level—even where a customer may only have a single ISP offering 100/20 Mbps speeds, almost all of those customers have at least one other option that creates competitive pressure.[30]

But the NPRM’s cited data is subject to greater criticism on other fronts. First, it cites to service availability data from 2021—a period during which ISPs were rapidly expanding their capacity to accommodate pandemic-era demands. The rate of this expansion was supplement by tens of billions of dollars of funding was made available through the American Rescue Plan Act (ARPA) to support broadband infrastructure investment, none of which is included in the cited data.[31] This data also doesn’t consider satellite options—Space X’s Starlink is now widely available throughout the United States[32] and Amazon’s Kupier project is beginning to put satellites into orbit.[33] And it excludes consideration of fixed wireless services, including both those commonly offered by WISPs and new 5G fixed wireless services offered by traditional wireless companies like T-Mobile and Verizon. The exclusion of fixed wireless services is notable given the pace at which consumers are embracing the technology as an alternative to traditional wireline service options.[34]

And, of course, the numbers cited by the NPRM exclude future broadband availability that will be facilitated by the Broadband Equity Access and Deployment (BEAD) program, which will increase broadband availability in currently un- and underserved areas.

The BEAD program draws attention to another concern about the proposed rules: over the past several years Congress has enacted numerous laws that support broadband investment but has not included net neutrality or open internet considerations in any of them. As will be discussed in the final section of these comments, the ultimate question for any federal agency action is whether it is authorized by Congress. The concept of net neutrality has been debated in Congress for more than a decade—it not conceivable that Congress was not aware of concerns such as those animating this NPRM as it developed the BEAD program. An inference can therefore be drawn from the fact that Congress, in allocating tens of billions of dollars in support for broadband investment, did not think it necessary to include provisions relating to network neutrality, even though doing so would settle a debate that has spanned four presidential administrations and produced substantial uncertainty within industry.

Indeed, a related inference might be drawn from Congress’s decision to entrust the National Telecommunications and Infrastructure Agency to oversee the BEAD program’s more-than $42 billion budget.[35] The BEAD program has many similarities to the Commission’s longstanding universal service programs, including raising many issues about which one would ordinarily assume the Commission as—and is viewed by Congress as having—substantial expertise. That Congress turned to another agency for the implementation of the largest broadband infrastructure program in the nation’s history should provide guidance on how broadly to interpret the Commission’s authority under the Communications Act to implement related program such as those proposed in the NPRM.[36]

III. First Amendment Considerations Cast Doubt on the Proposed Rules

The proposed rules raise at least three sets of concerns under the First Amendment. First, whether imposition of common carriage obligations is problematic under First Amendment principles. Second, whether the specific proposed rules are problematic under First Amendment principles, even absent reclassification. And third, the relevance of cases currently pending before the Supreme Court to the proposed rules.

The last of these concerns can be addressed with brevity: NetChoice v. Paxton is currently pending before the Supreme Court, with an opinion expected this term.[37] This case asks the Court to consider whether laws that limit social media platforms’ content-moderation practices, imposing common-carriage-like obligations upon them, comply with the First Amendment. The laws that the Court is reviewing bear similarity to the rules proposed by the Commission and there is a high likelihood that the Court’s decision with have significant bearing on these proposed rules. It would be imprudent to adopt any rules on this topic until the Supreme Court issues its opinion in NetChoice.

The NetChoice case also illustrates the next point to consider: whether imposition of common carriage obligations is problematic under the First Amendment. There is a long history of efforts to impose such obligations on communications platforms, from the social media at issue in NetChoice, to radio and television broadcasters, news papers, cable networks, and telephone networks.[38] But it turns out that the idea of “imposing common carriage obligations” gets the question backwards. The most universally accepted definition of common carriage turns on whether the firm eschews exercising editorial discretion over the content it carries and instead holds itself out as serving all members of the public without engaging in individualized bargaining.[39] In other words, the question is not whether government can impose common carriage obligations—it is whether a firm conducts its business in the style of a common carrier. Doing so can create an obligation to honor the obligations of common carriage (which may come with regulatory benefits). But a firm can avoid those obligations by engaging in individualized bargaining. This criterion constitutes the central consideration in all leading discussions of common carriage.[40]

This leads to the curious conclusion that the Commission’s proposal to classify BIAS as a Title II telecommunications service—a common carriage service—would only have effect to the extent that a BIAS provider elects to hold itself out as offering a common carriage service. This conclusion is reflected in the exchange that took place during the D.C. Circuit’s decision not to rehear the decision upholding the 2015 Open Internet Order en banc. When then-Judge Kavanaugh objected that classifying ISPs as common carriers impermissibly abridged their editorial discretion,[41] the authors of the majority opinion countered, explaining that “[w]hen a broadband provider holds itself out as giving customers neutral, indiscriminate access to web content of their own choosing, the First Amendment poses no obstacle to holding the provider to its representation.”[42] The 2015 Open Internet Order did not implicate the First Amendment because it only purported to regulate services over which providers exercised no editorial discretion.[43] This discussion implicitly recognized that speech over which providers exercise editorial control is protected by the First Amendment. If that were not the case, the fact that the 2015 Open Internet Order affected only speech over which providers exercised no editorial discretion would have been completely unresponsive to the concerns raised by then-Judge Kavanaugh. As described by the majority, this is a “if you say it, do it” theory, nothing more.[44]

Indeed, the initial Circuit Court opinion reached a similar conclusion—and noted the 2015 Open Internet Order recognized as much as well:

If a broadband provider nonetheless were to choose to exercise editorial discretion—for instance, by picking a limited set of websites to carry and offering that service as a curated internet experience—it might then qualify as a First Amendment speaker. But the Order itself excludes such providers from the rules. . . . Providers that may opt to exercise editorial discretion—for instance, by offering access only to a limited segment of websites specifically catered to certain content—would not offer a standardized service that can reach “substantially all” endpoints. The rules therefore would not apply to such providers, as the FCC has affirmed.[45]

The perverse incentive this creates bears emphasis. It is clear from the history of the Commission’s Open Internet proceedings that BIAS providers do not want to be subject to Title II regulation. These providers might today hold themselves out as providing services that might be considered by consumers to be common carriage services. If these providers want to avoid the obligations of Title II regulation, they would be able to do so by reducing the scope of their services to the point that they cannot be seen as holding themselves out as common carriers. The First Amendment affords no path for the Commission to compel ISPs to do otherwise.

A question still remains whether the specific rules proposed in the NPRM are problematic under First Amendment principles, even absent reclassification. In Turner, the Supreme Court upheld “must-carry” rules for cable networks, which are not common carriers, even where those networks have some claim to editorial discretion.[46] The result in Turner, however, turned on the “gatekeeper” or “bottleneck” control resulting from “the fact that there could only be one cable connection to any home” that places the cable operator in a position to block any other content providers from gaining access to subscribers.[47] In so holding, the Court emphasized the physical (rather than economic) nature of this consideration by contrasting cable with newspapers, which, “no matter how secure its local monopoly, does not possess the power to obstruct readers’ access to other competing publications.”[48] In an era of multimodal communications, where households routinely have connections to telephone, cable, and fiber optic networks, as well is fixed wireless, mobile wireless, and satellite connectivity options, makes clear the inapplicability of this rationale to platforms that lack control over an exclusive physical connection precludes applying this rationale to ISPs.

IV. The Major Questions Doctrine Casts Doubt on the Proposed Rules

The NPRM asks dozens of questions over several paragraphs relating to the Major Questions Doctrine.[49] First formally recognized by the Supreme Court in 2022, this doctrine explains that the Court “expect[s] Congress to speak clearly if it wishes to assign to an agency decisions of vast ‘economic and political significance.’”[50] Although only first recognized by the Court in 2022, the Major Questions Doctrine has developed over a span of many decades, through both Supreme Court cases and those decided by lower courts.[51] Importantly, nothing in these cases suggests that an agency’s recognition that its decisions may raise major questions renders those decisions any less major. Rather, the fact that an agency feels it is necessary to ask whether its decisions raise major questions suggests that those questions may well be major. This alone should give the agency pause about taking such decisions—especially in an era of intense judicial scrutiny of agency action it would a curious decision for any agency instead seek to structure its decisions so as to avoid the appearance of their having vast economic or political significance.

The vast significance of the proposed rules cannot be overstated—though the NPRM seems notably to attempt to understate it. Discussing broadband in 2015, former Chair Tom Wheeler described the Internet as “the most powerful network in the history of mankind.”[52] This was echoed in the 2015 Open Internet Order. The first sentence of the 2015 Order asserted that “[t]he open Internet drives the American economy and serves, every day, as a critical tool for America’s citizens to conduct commerce, communicate, educate, entertain, and engage in the world around them.”[53] Similarly, the NPRM for that Order began with: “The Internet is America’s most important platform for economic growth, innovation, competition, [ and] free expression . . . . [It] has been, and remains to date, the preeminent 21st century engine for innovation and the economic and social benefits that follow.”[54]

And, as made clear in the current NPRM, this importance has only been amplified since the beginning of the COVID-19 pandemic: “In the time since the RIF Order, propelled by the COVID-19 pandemic, BIAS has become even more essential to consumers for work, health, education, community, and everyday life.”[55] As stated by Chair Rosenworcel, “broadband is no longer nice-to-have; it’s need-to-have for everyone, everywhere. Broadband is an essential service”[56] Commissioner Starks similarly states that “there is simply no way to overstate broadband’s impact on the lives of individual Americans.”[57] And Commissioner Gomez states “El acceso a internet de banda ancha no solo es una herramienta vital para la educación, la atención de salud y para comunicarnos con nuestros seres queridos. También es un conducto de crítica importancia, esencial para la vida moderna.”[58]

In the NPRM, the Commission points to the D.C. Circuit’s opinion in the previous Open Internet Order case, as well as to Brand X itself, to suggest that the Brand X opinion settles the question of whether the Commission has authority to adopt these rules.[59] As an initial matter, that question simply was not before the Court in Brand X. In Brand X, the Court considered only whether the FCC could decide the classification under the Chevron after a lower court had adopted a conflicting interpretation of an ambiguous statute.[60] The question of whether Commission had authority to regulate cable broadband Internet service as a Title II telecommunications service was not a contested issue—only whether it could instead interpret ambiguity in the Communications Act to instead regulate it as a Title I information service. Similarly, the Major Questions Doctrine had not been expressly recognized by the Supreme Court at the time of the D.C. Circuit’s 2016 opinion. While it was “in the air,” it was a controversial doctrine—one that the Supreme Court had yet to expressly embrace. But the Court has done so now.

In the D.C. Circuit’s denial of en banc review, the judges Srinivasan and Tatel (the authors of the majority opinion under review) offers a more fulsome consideration of the major questions issue[61]—prompted by the dissenting opinion of then-judge Kavanaugh.[62] But in light of the Supreme Court’s own more fulsome embrace of the doctrine since that opinion, this analysis is unconvincing. In addition to over-relying on Brand X as having decided the matter, the majority makes another conceptually uncertain assumption: even if the Commission could have regulated Internet services at the time of Brand X, that doesn’t necessarily mean that the Commission has such authority today. The economic and political significance of those services is vastly different today than it was then. The Commission disclaimed such authority for more than 15 years, during which time the social, economic, technical, and political understandings of the Internet changed fundamentally from what they were at the time of Brand X. As discussed in the following paragraphs, this consideration is relevant to the Major Questions Doctrine analysis but goes unrecognized by the D.C. Circuit opinion.

While Brand X does not support the contention that classifying BIAS as a Title II does not present major question, it does briefly draw attention to an important argument that it does present a major question: respondents in Brand X raised the concern that “the Commission’s construction [of cable Internet as a Title I information service] is unreasonable because it allows any communications provider to ‘evade’ common-carrier regulation by the expedient of bundling information service with telecommunications.”[63] The Court rejected the premise of this argument, so did not decide whether it would, in fact, be an unreasonable construction.[64] But the same issue runs through the proposed rules: that BIAS providers need not hold out their offerings as common carriage services, so could “evade” Title II regulation through a simple expedient. Unlike in Brand X, where the Court rejected the premise that such evasion was possible, here the D.C. Circuit has found that the First Amendment compels such a result and the Commission has conceded that the rules necessarily allow it.[65]

This is a facially absurd result that demonstrates the fundamental mismatch between Title II regulation at BIAS service. Title II is a pervasive regulatory regime designed to regulate one of the foundational, utility-style, natural monopoly, industries of the 20^th century. It was designed for a technology to which editorial discretion was largely viewed as inapplicable.[66] And it was part of a regulatory quid pro quo that gave telephone carriers a privileged regulatory position in exchange for offering service on a common-carriage basis. As we approached the 21^st century, Title II was substantially amended through the 1996 Telecommunications Act to be a pervasively deregulatory statute that would phase out regulate in favor of competition and did so largely in response to the development of networks that offered advanced communications capabilities that were increasingly removed from the transparent, point-to-point, communications model of the 20^th century telephone network.[67]

It is little surprise that the Commission’s efforts today to reregulate modern communications networks under an expressly deregulatory act runs into problems such as the possibility for evasion. Similarly, it is unsurprising that it required tailoring of the Title II regulatory regime through extensive use of forbearance—another hallmark of a regulatory decision that presents major questions.[68]

The idea that Congress intends the Commission to be the primary regulator for BIAS services faces even greater headwind following the pandemic. Congress—which has longstanding awareness of the net neutrality debates—put in place significant communications-related regulations with the IIJA. This included the BEAD program, the Affordable Connectivity Program, and laying the groundwork for the Commission’s recently proposed Digital Discrimination rules.[69] This legislation, adopted during a period of single-party control of Congress and the White House, could have easily clarified the Commission’s regulatory authority over BIAS. Instead, it assigned primary authority for the federal government’s flagship broadband infrastructure program to another agency and assigned to the Commission new authority to adopt more limited broadband discrimination rules to ensure equitable access to that infrastructure.

These circumstances bear resemblance to the circumstances discussed by the Court in FDA v. Brown & Williamson Tobacco Corp, one of the precursor cases to the Major Questions Doctrine.[70] In Brown & Williamson, the Court held that the FDA did not have authority to regulate tobacco, a drug, under its statutory authority to regulate drugs. In reaching this conclusion, the Court recognized among other things that Congress knew the FDA did not have a clear claim to regulate tobacco and had adopted alternative regulatory regimes relating to the regulation of tobacco without giving the FDA regulatory authority over the matter.[71] It also recognized that “it is hardly conceivable that Congress—and in this setting, any Member of Congress—was not abundantly aware of what was going on.”[72] Here, as there, if Congress intended the FCC to exercise the vast authority that the Commission would claim in the NPRM, Congress was aware of the uncertainty over the Commission’s authority and could readily have clarified the matter.

Conclusion

The defining feature of the Safeguarding and Securing the Open Internet NPRM is its proposed reclassification of BIAS as a Title II telecommunication service. By choosing this approach, the Commission is needlessly steering into both legal uncertainty and political controversy. Reclassification will be challenged in court; these challenges have a high likelihood of success and, in the best case, will yield years of uncertainty for industry, consumers, and the Commission. Indeed, cases currently pending before the Supreme Court could force reconsideration of the NPRM before the proposed rules could even be finalized.

This is a curious approach to take, given that ISPs can, in effect, “opt out” of reclassification by offering an Internet experience that does not purport to be a common carriage service. This would be an unfortunate outcome that would lead to consumer confusion and possibly to degraded user experiences. It is a perverse approach to regulation, imposing burdensome rules that can be avoided by degrading the quality of service offered to consumers. And this demonstrates the extent to which the proposed reclassification is a perversion of the purposes of Title II—and to which it therefore is a decision that raises major questions.

And the approach is all the more curious given that there is little demonstrable need for the rules in the first place. But if the rules are to be adopted, the Commission could do so without reclassification. Were the Commission to follow the roadmap offered by the D.C. Circuit in Verizon, it is entirely likely that industry would acquiesce to the rules. That is the path the Commission should take, if it is to go down this path at all.

[1]    Safeguarding and Securing the Open Internet, WC Docket No. 23-230, Notice of Proposed Rulemaking, 88 Fed. Reg. 76,048 (Nov. 3, 2023) (NPRM).

[2]    Further detail on many of the topics discussed in these comments can be found in our prior publications. See, e.g., [[past comments, briefs, articles]].

[3]    Consolidated Appropriations Act, 2021, Pub. L. No. 116-260, § 904 (2020).

[4]    American Rescue Plan Act of 2021, Pub. L. No. 117-2, § 7402 (2021) (ARPA).

[5]    Infrastructure Investment and Jobs Act, Pub. L. No. 117-58, § 60502 (2021) (IIJA)

[6]    Id. § 60101, et seq.

[7]    Id. § 60102.

[8]    See Tom Wheeler, Finding the Best Path Forward to Protect the Open Internet (April 29, 2014), available at https://www.fcc.gov/news-events/blog/2014/04/29/finding-best-path-forward-protect-open-internet (“In its Verizon v. FCC decision the D.C. Circuit laid out a blueprint for how the FCC could use Section 706 of the Telecommunications Act of 1996 to create Open Internet rules that would stick. I have repeatedly stated that I viewed the court’s ruling as an invitation that I intended to accept.”).

[9]    See Part IV.

[10]   Nat’l Cable & Telecomms. Ass’n v. Brand X, 545 U.S. 967 (2005) (Brand X).

[11]   See Part I.

[12]   See Part III.

[13]   Brand X at 989 (“Cable companies in the broadband Internet service business ‘offer’ consumers an information service in the form of Internet access and they do so “via telecommunications.”).

[14]   Id. At 990 (“The question, then, is whether the transmission component of cable modem service is sufficiently integrated with the finished service to make it reasonable to describe the two as a single, integrated offering”). More precisely, the Court was considering whether this interpretation, made by the FCC, can be sustained under the Chevron doctrine. Id. at 989 (“This construction passes Chevron’s first step.”); id. at 997 (“We also conclude that the Commission’s construction was ‘a reasonable policy choice for the Commission to make’ at Chevron’s second step.”).

[15]   NPRM, paras. 11, 75–77 (citing U.S.Telecom Ass’n v. FCC, 825 F.3d 674(D.C. Cir. 2016) (USTA))

[16]   USTA at 705.

[17]   Id. at 766, n.8 (Williams, J, dissenting).

[18]   Id. at 748.

[19]   Brand X at 991.

[20]   Id. at 1012–13 (Scalia, J, dissenting).

[21]   Id. at 999, n.3.

[22]   See Reply Comments of Christopher S. Yoo, WC Docket No. 17-108.

[23]   A Non-Queue-Building Per-Hop Behavior (NQB PHB) for Differentiated Services, IETF Transport Area Working Group Draft (Oct. 24, 2022), available at https://www.ietf.org/archive/id/draft-ietf-tsvwg-nqb-14.html.

[24]   Low Latency, Low Loss, Scalable Throughput (L4S) Internet Service: Architecture, IETF Transport Area Working Group Draft  (July 27, 2022), available at https://www.ietf.org/archive/id/draft-ietf-tsvwg-l4s-arch-19.html. See also Mitchell Clark, The Quiet Plan to Make the Internet Feel Faster, The Verge (Dec. 9, 2023).

[25]   See, e.g., Joan Feigenbaum et al, A BGP-based Mechanism for Lowest-Cost Routing, 18 Distributed Computing 61 (2005).

[26]   Geoff Huston, Looking at Centrality in the DNS,APNIC Blog (Nov. 22, 2022), available at  https://blog.apnic.net/2022/11/22/looking-at-centrality-in-the-dns/ (“the use of ISP-provided recursive resolution occurs for between 65% to 80% of users . . . known open resolvers have a 20% market share”).

[27]   Id.

[28]   Firefox DNS-over-HTTPS, Mozilla Support, available at https://support.mozilla.org/en-US/kb/firefox-dns-over-https (“When DoH is enabled, Firefox by default directs DoH queries to DNS servers that are operated by a trusted partner . . . .”) (visited Dec. 13, 2023).

[29]   NPRM, paras 126, 128.

[30]   Importantly, and contrary to the longstanding policy approach of defining performance metrics that focus on increasingly higher bandwidth (Mbps targets), “It has been demonstrated that, once access network bit rates reach levels now common in the developed world, increasing link capacity offers diminishing returns if latency (delay) is not addressed.” Low Latency, Low Loss, Scalable Throughput (L4S) Internet Service: Architecture, supra note 24. Implementation of new technologies, such as L4S, could increase competition between existing networks than at lower cost than building out entirely new, increasingly higher-speed, infrastructure.

[31]   See Anna Read & Kelly Wert, How States Are Using Pandemic Relief Funds to Boost Broadband Access, Pew (Dec. 6, 2021), available at https://www.pewtrusts.org/en/research-and-analysis/articles/2021/12/06/how-states-are-using-pandemic-relief-funds-to-boost-broadband-access.

[32]   Starlink Availability Map, available at https://www.starlink.com/map.

[33]   See Joey Roulette, Amazon’s Prototype Kuiper Satellites Operating Successfully, Reuters (Nov. 16, 2023), available at https://www.reuters.com/technology/amazons-prototype-kuiper-satellites-operate-successfully-2023-11-16/.

[34]   See Mike Dano, T-Mobile Exceeds in Q3, Talks Broadband Strategy, Light Reading (Oct. 25, 2023), available at https://www.lightreading.com/fixed-wireless-access/t-mobile-exceeds-in-q3-talks-broadband-strategy (“T-Mobile reported a total of 557,000 new fixed wireless access (FWA) customers, a figure above most analyst expectations and slightly ahead of the company’s recent quarterly pace.”); Mike Dano, FWA Captures 90% of All New US Customers, Pleasing Around 90% of Them, Light Reading (March 6, 2023), available at https://www.lightreading.com/fixed-wireless-access/fwa-captures-90-of-all-new-us-customers-pleasing-around-90-of-them (“fixed wireless services accounted for 90% of all net broadband customer additions in the US during 2022 . . . . 90% rated their service as ‘good enough.’”).

[35]   IIJA § 60502(b)(2).

[36]   See infra, Part IV.

[37]   Moody v. NetChoice, LLC; NetChoice, LLC v. Moody; NetChoice, LLC v. Paxton, Nos. 22-277, 22-393 and 22-555 (U.S. Sup. Ct.).

[38]   See generally Christopher Yoo, Free Speech and the Myth of the Internet as an Unintermediated Experience, 78 Geo. Wash. L. rev. 697, Part II (2010) (surveying regulation of various media technologies).

[39]   See Biden v. Knight First Amendment Inst., 141 S. Ct. 1220, 1222 (2021) (Thomas, J., concurring).

[40]   See FCC v. Midwest Video Corp., 440 U.S. 689, 701 (1979); U.S. Telecom Ass’n v. FCC, 825 F.3d 674, 740 (D.C. Cir. 2016); Verizon v. FCC, 740 F.3d 623, 651 (D.C. Cir. 2014); Cellco P’ship v. FCC, 700 F.3d 534, 548 (D.C. Cir. 2012); NARUC II, 533 F.2d at 608; NARUC I, 525 F.2d at 641. Congress, courts, and agencies have applied the same formulation in a wide variety of contexts. See 15 U.S.C. § 375(3); 46 U.S.C. § 40102(7)(A); 40 C.F.R. § 202.10(b); Edwards v. Pac. Fruit Express Co., 390 U.S. 538, 540 (1968); Woolsey v. Nat’l Transp. Safety Bd., 993 F.2d 516 524 n.2. (5th Cir. 1993); Flytenow, Inc. v. FAA, 808 F.3d 882, 887–88 (D.C. Cir. 2015); Nichimen Co. v. M. V. Farland, 462 F.2d 319, 326 (2d Cir. 1972); Kelly v. Gen. Elec. Co., 110 F. Supp. 4, 6 (E.D. Pa.), aff’d, 204 F.2d 692 (3d Cir. 1953).

[41]   U.S. Telecom Ass’n, 855 F.3d at 484–89 (Kavanaugh, J., dissenting from the denial of rehearing en banc)

[42]   Id. at 392 (Srinivasan, J., joined by Tatel, J., concurring in the denial of the petition for rehearing en banc).

[43]   Id. at 388–89.

[44]   Id. at 392.

[45]   USTA at 743.

[46]   Turner Broad. Sys., Inc. v. FCC (Turner I), 512 U.S. 622, 656–57 (1994).

[47]   Id. at 656.

[48]   Id.

[49]   NPRM, paras 81–84.

[50]   West Virginia v. EPA, 142 S. Ct. 2587 (2022) (quoting Utility Air Regulatory Group v. EPA, 573 U.S. 302, 324 (2014)).

[51]   Id. at 2609 (“major questions doctrine label . . . took hold because it refers to an identifiable body of law that has developed over a series of significant cases all addressing a particular and recurring problem: agencies asserting highly consequential power beyond what Congress could reasonably be understood to have granted. Scholars and jurists have recognized the common threads between those decisions. So have we.” (citing cases dating to MCI Telecommunications Corp. v. American Telephone & Telegraph Co., 512 U.S. 218 (1994) (MCI)).

[52]   See Remarks of FCC Chairman Tom Wheeler, Silicon Flatirons Center (Feb. 9, 2015), available at http://transition.fcc.gov/Daily_Releases/Daily_Business/2015/db0209/DOC-331943A1.pdf; see also of same, available at https://www.youtube.com/watch?v=vHsHkKpxVkQ (in which Chairman Wheeler was more emphatic than in his prepared remarks); Brian Fung, FCC chairman warns: The GOP’s net neutrality bill could jeopardize broadband’s ‘vast future’, Washington Post (Jan. 29, 2015), available at http://www.washingtonpost.com/blogs/the-switch/wp/2015/01/29/fcc-chairman-warns-that-republican-bill-couldjeopardize-broadbands-vast-future/.

[53]   Protecting and Promoting the Open Internet, GN Docket No. 14-28, Report and Order on Remand, Declaratory Ruling, and Order, 80 Fed. Red. 19,737 (Apr. 15, 2015).

[54]   Protecting and Promoting the Open Internet, GN Docket No. 14-28, Notice of Proposed Rulemaking, 79 Fed. Reg. 37,448 (July 1, 2014).

[55]   NPRM, para 16.

[56]   NPRM, Statement of Chairwoman Jessica Rosenworcel.

[57]   NPRM, Statement of Commissioner Geoffrey Starks.

[58]   NPRM, Statement of Commissioner Anna M. Gomez.

[59]   NPRM, para 81 (“In the USTA decision, the D.C. Circuit reasoned that Brand X conclusively held that the Commission has the authority to determine the proper statutory classification of BIAS and that its determinations are entitled to deference, and so there is no need to consult the major questions doctrine here.”).

[60]   Brand X, at 974 (“We must decide whether [the FCC’s] conclusion [that cable companies that sell broadband Internet service do not provide ‘telecommunications service’ as the Communications Act defines that term] is a lawful construction of the Communications Act under Chevron U. S. A. Inc. v. Natural Resources Defense Council, Inc. and the Administrative Procedure Act. We hold that it is.”).

[61]   U.S. Telecom Ass’n, 855 F.3d at 385–88.

[62]   U.S. Telecom Ass’n, 855 F.3d at 422–26 (Kavanaugh, J., dissenting).

[63]   Brand X, at 997.

[64]   Id. (“We need not decide whether a construction that resulted in these consequences would be unreasonable because we do not believe that these results follow from the construction the Commission adopted.”).

[65]   See supra, Part III.

[66]   But see Christopher Yoo, Free Speech and the Myth of the Internet as an Unintermediated Experience, 78 Geo. Wash. L. rev. 697, 752–57 (2010) (discussing cases starting in the 1980s that began to recognize the applicability of editorial discretion concepts to telephone carriers).

[67]   See Telecommunications Act of 1996, Pub. L. No. 104-104 (1996) (“An act to promote competition and reduce regulation in order to secure lower prices and higher quality services for American telecommunications consumers and encourage the rapid deployment of new telecommunications technologies.”).

[68]   See U.S. Telecom Ass’n, 855 F.3d at 404–09 (Brown, J., dissenting from the denial of rehearing en banc) (citing MCI and Utility Air Regulatory Group v. EPA in discussing the problematic use of the Commission’s forbearance authority to “tailor” Title II to fit the needs of BIAS regulation).

[69]   See supra, notes 3–5.

[70]   FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120 (2000).

[71]   Id. at 155-156.

[72]   Id. at 156.