Contents

Introduction and Summary 1

The Record Confirms That Section 60506 Empowers the Commission To Target Disparate Treatment, Not Disparate Impact 1

The Record Does Not Show Digital Discrimination To Be a Real-World Problem. 3

Heavy-Handed Rules Will Undermine Efforts To Close the Digital Divide. 4

Conclusion. 4

Endnotes 4

Introduction and Summary

ITIF appreciates this opportunity to reply to comments in the Federal Communications Commission’s (“FCC” or “Commission”) Notice of Proposed Rulemaking (“NPRM” or “Notice”) on the implementation of the provisions of the Infrastructure and Jobs Act (“IIJA”) related to purported “digital discrimination.”[1]

The Record Confirms That Section 60506 Empowers the Commission To Target Disparate Treatment, Not Disparate Impact

Though many commenters attempt to squeeze a disparate impact definition out of the language of the statute, none are successful. As ITIF explained in its initial comments, the Supreme Court requires “specific language in a legislative prohibition of discrimination broadens its scope” to include disparate impact claims. Proponents of a disparate impact standard for Section 60506 do not root their arguments in the text of the operative provision, subsection (b)(1), but in appeals to other parts of the statute or to extra-statutory opinion.

Many supporters of a disparate impact definition of digital discrimination base their arguments on the absence of statutory language. The American Library Association, for example, claims that the Commission should “should adopt rules that encompass disparate impact claims because the statute does not specify that intent is a required element of digital discrimination.”[2] The County of Santa Clara likewise points to the lack of statutory language requiring “malicious intent” as justifying disparate impact liability.[3] But citing the absence of text to justify the extreme step of imposing liability for disparate impacts is the opposite of the Inclusive Communities standard. Moreover, the statute does specify that digital discrimination is an act “based on” protected characteristics, which does entail a disparate treatment definition.[4]

Other commenters who advocate a disparate impact definition do so on the basis of their own policy preferences without regard to the statutory text. The City of Minneapolis Department of Civil Rights provides a typical example. It supports a disparate impact standard because it would “give the FCC flexibility it needs to address the widest possible range of complaints.”[5] This standard is as nonsensical as it is non-textual. If the Commission were able to simply wield whatever standard it found most “flexible,” there would be no need for statutory grants of authority. But, on the contrary, the Commission is bound by statutes and may simply ignore them to do what it deems convenient.

Other disparate-impact proponents look to other provisions of the statute. The Lawyers’ Committee for Civil Rights Under Law claims that disparate impact liability is necessitated by the fact that “the text is aimed at achieving an outcome.”[6] But all statutes do that. The fact that the FCC’s rules will be a means to an end does not empower the Commission to do whatever it wants as long as it has the same end in mind. The Lawyers’ Committee goes on to import subsection 50506(a) into the operative provision of 60506(b). Besides straining the meaning of the text, this move is also foreign to the Inclusive Communities test in which the language about consequences was in the prohibition itself rather than in expressions of intent in other provisions.[7]

The Lawyers’ Committee repeats its argument that “the use of the ‘technical and economic feasibility’ terms in Section 60506 is strong evidence for disparate impact coverage” since it would be absurd to think that “a business could excuse purposeful racial animus on economic grounds.”[8] As ITIF explained in its initial comments, this argument misconstrues the text.[9] The statute requires the Commission to take “into account the issues of technical and economic feasibility.” Technical and economic feasibility are thus considerations for the Commission in making its rules, not a statutorily created defense to intentional discrimination.[10] To be sure, if the Commission adopted a disparate impact definition, it would have to allow “business necessity” defenses, but that is because it is an attribute of antidiscrimination law, not because of the economic feasibility language in the statute.

Commenters who actually look to the text of 60506(b) come to the conclusion that it does not permit disparate impact claims.[11] The Commission must be guided by its governing statutes and not grasp at more power than Congress has given it.

The Record Does Not Show Digital Discrimination To Be a Real-World Problem

In addition to the lack of statutory authority for sweeping regulation of broadband deployment, there is also little evidence in the record to suggest that there is even a discrimination problem in need of a crackdown on ISPs. If digital discrimination were a problem necessitating the draconian measures some commenters propose, strong evidence should be forthcoming. Instead, no new studies were submitted that even allege, let alone prove, that ISPs engage in discriminatory conduct. The purported evidence of discrimination in the record relies on innuendo and bare allegation rather than rigorous causal analysis. Commenters point to alleged “underinvestment by ISPs in traditionally redlined communities” or complaints filed by residents who were dissatisfied with their broadband service.[12] The Commission would be right to consider policies to enhance the broadband opportunities of all Americans including those who have been historically disadvantaged. But a kneejerk reaction to construct a new regulatory apparatus would go far beyond what evidence warrants, and the Commission would be on shaky ground if it does so.

Moreover, data in the record find the opposite conclusion. Updated data from the Commission’s former Chief Economist Glenn Woroch finds that broadband deployment does not “systematically and disproportionately under-serve lower-income or non-white households.”[13] A Cartesian study of cable providers likewise found “no evidence” of discrimination “in the availability of high-speed services within its service footprint, based on either income, race or ethnicity.”[14] The Phoenix Center conducted another economic analysis which found “no meaningful evidence of digital discrimination in either race or income for fiber deployments or for download speeds.”[15] Each of these findings is consistent with an earlier ITIF study that found ISPs do not engage in “digital redlining.”[16]

The National Digital Inclusion Alliance (“NDIA”) takes issue with ITIF’s findings.[17] NDIA seems to agree with ITIF’s premise that it is unlikely that ISPs will discriminate based on irrelevant factors since that would entail the ISP not seeking to maximize profits.[18] Nevertheless, NDIA seeks expansive liability under Section 60506 because “[t]he goals of ‘universal service’ and ‘maximize profit’ are irreconcilable.” This proves too much. What NDIA’s statement really advocates is the radical step of defining digital discrimination as “ISPs not operating at a loss” which would transform the present antidiscrimination statute into a national deployment mandate and back-door nationalization of a massive industry. Such a pronouncement would be one of great “magnitude and consequence” for which the Commission does not have a “clear delegation” from Congress; thus, it would unconstitutionally exceed the Commission’s authority.[19] Moreover, NDIA’s attempt to oppose private broadband and universal service undermines the latter as well as the former. It is certainly true that the FCC should advance the goal of universal service, but that should occur in collaboration with the private, for-profit broadband investments that have powered the success of the American Internet ecosystem.[20]

And, on the other side of the equation, it is untrue that government-owned networks are non-discriminatory. On the contrary, many municipal broadband efforts have often cherrypicked low-cost, high-return areas to serve at the expense of lower-return ones, a pattern that results in disparities along racial and income lines. Chattanooga’s Electric Power Board focuses on central areas, and even then it has been unable to survive without ongoing subsidies.[21] Ammon, Idaho’s government network requires consumers to pay upfront costs of $3,200-3,600.[22] Such a measure disproportionately impacts lower-income neighborhoods. In many cases, municipal networks are only deployed in disproportionately wealthy areas. Concord, Massachusetts’s Concord Light Broadband serves an area whose median household income is more that $80,000 higher than the statewide median income.[23] Unlike private providers that serve much broader geographic areas in the Boston region, the Concord government by definition “cream skims” because it does not have to serve lower-return geographic areas. If the Commission seeks to eliminate these kinds of income disparities, it should view government-owned networks as part of the problem, not the solution.

Heavy-Handed Rules Will Undermine Efforts To Close the Digital Divide

ITIF consistently supports efforts to close the digital divide, and the Commission must be careful that its rules do not counterproductively reduce the rate at which that divide closes. The record is consistent with ITIF’s findings that adoption, not deployment, is the main remaining barrier to closing the digital divide. Any remaining geographically driven divide will be addressed with funding from the BEAD and other subsidy programs. Even if the Commission undertook to identify all areas of unequal deployment, its evaluation would quickly become obsolete after these unprecedented levels of funding connect the previously unserved. Adding a liability risk for benign deployment processes, such as upgrading parts of an ISP’s footprint over time rather than all at once, will reduce the effectiveness of those programs and delay broadband access for those who need it most.

Indeed, even many supporters of more sweeping rules lodge criticisms that are actually manifestations of adoption barriers, not deployment decisions. Calls for the Commission to have the “widest possible range of complaints” manifests the diversity of reasons individuals have for not being online. If the problem were really a lack of deployment by ISPs, the range of complaints would be very narrow. But adoption is multifaceted problem linked to many factors outside the control of ISPs.

It is true that unequal usage of broadband persists, but the Commission should tune its rules to address the real causes of unequal access rather than imposing a punitive regulatory overhang that does not address the root problems. Focusing on access to devices, individual affordability (such as the Affordable Connectivity Program), and adoption efforts (like digital navigators) will yield far more equal access by treating the cause of unequal access rather than imposing punitive liability on parties who did not proximately cause it.

Conclusion

The Commission’s rules should adhere to the text of the statute and use evidence-based policies to promote the ends of universal broadband access.

Endnotes