Restoring Internet Freedom

FEDERAL COMMUNICATIONS COMMISSION

47 CFR Parts 1, 8, and 20

[WC Docket No. 17-108; FCC 17-166]

Restoring Internet Freedom

AGENCY:

Federal Communications Commission.

ACTION:

Final rule.

SUMMARY:

In this document, the Federal Communications Commission (Commission) returns to the light-touch regulatory scheme that enabled the internet to develop and thrive for nearly two decades. The Commission restores the classification of broadband internet access service as a lightly-regulated information service and reinstates the private mobile service classification of mobile broadband internet access service. The

Restoring Internet Freedom Order

requires internet service providers (ISPs) to disclose information about their network management practices, performance characteristics, and commercial terms of service. Finding that transparency is sufficient to protect the openness of the internet and that conduct rules have greater costs than benefits, the Order eliminates the conduct rules imposed by the

Title II Order.

DATES:

Effective date:

April 23, 2018, except for amendatory instructions 2, 3, 5, 6, and 8, which are delayed as follows. The FCC will publish a document in the

Federal Register

announcing the effective date(s) of the delayed amendatory instructions, which are contingent on OMB approval of the modified information collection requirements in 47 CFR 8.1 (amendatory instruction 5). The Declaratory Ruling, Report and Order, and Order will also be effective upon the date announced in that same document.

FOR FURTHER INFORMATION CONTACT:

Ramesh Nagarajan, Competition Policy Division, Wireline Competition Bureau, at (202) 418-2582,

ramesh.nagarajan@fcc.gov.

For additional information concerning the Paperwork Reduction Act information collection requirements contained in this document, send an email to

PRA@fcc.gov

or contact Nicole Ongele at (202) 418-2991.

SUPPLEMENTARY INFORMATION:

This is a summary of the Commission's Declaratory Ruling, Report and Order, and Order (“

Restoring Internet Freedom Order”

) in WC Docket No. 17-108, adopted on December 14, 2017 and released on January 4, 2018. The full text of this document is available at

https://apps.fcc.gov/edocs_public/attachmatch/FCC-17-166A1.pdf.

The full text is also available for public inspection during regular business hours in the FCC Reference Information Center, Portals II, 445 12th Street SW, Room CY-A257, Washington, DC 20554. To request materials in accessible formats for people with disabilities (

e.g.,

braille, large print, electronic files, audio format, etc.) or to request reasonable accommodations (

e.g.,

accessible format documents, sign language interpreters, CART, etc.), send an email to

fcc504@fcc.gov

or call the Consumer & Governmental Affairs Bureau at (202) 418-0530 (voice) or (202) 418-0432 (TTY). The language following the

DATES

caption of this preamble is provided to ensure compliance with 1 CFR 18.17.

Synopsis

In this Declaratory Ruling, Report and Order, and Order, the Commission restores the light-touch regulatory scheme that fostered the internet's growth, openness, and freedom. Through these actions, we advance our critical work to promote broadband deployment in rural America and infrastructure investment throughout the nation, brighten the future of innovation both within networks and at their edge, and move closer to the goal of eliminating the digital divide.

I. Ending Public-Utility Regulation of the Internet

1. We reinstate the information service classification of broadband internet access service, consistent with the Supreme Court's holding in

Brand X.

Based on the record before us, we conclude that the best reading of the relevant definitional provisions of the Act supports classifying broadband internet access service as an information service. Having determined that broadband internet access service, regardless of whether offered using fixed or mobile technologies, is an information service under the Act, we also conclude that as an information service, mobile broadband internet access service should not be classified as a commercial mobile service or its functional equivalent. We find that it is well within our legal authority to classify broadband internet access service as an information service, and reclassification also comports with applicable law governing agency decisions to change course. While we find our legal analysis sufficient on its own to support an information service classification of broadband internet access service, strong public policy considerations further weigh in favor of an information service classification. Below, we find that economic theory, empirical data, and even anecdotal evidence also counsel against imposing public-utility style regulation on ISPs. The broader internet ecosystem thrived under the light-touch regulatory treatment of Title I, with massive investment and innovation by both ISPs and edge providers, leading to previously unimagined technological developments and services. We conclude that a return to Title I classification will facilitate critical broadband investment and innovation by removing regulatory uncertainty and lowering compliance costs.

A. Reinstating the Information Service Classification of Broadband Internet Access Service

1. Scope

2. We continue to define “broadband internet access service” as a mass-market retail service by wire or radio that provides the capability to transmit data to and receive data from all or substantially all internet endpoints, including any capabilities that are incidental to and enable the operation of the communications service, but excluding dial-up internet access service. By mass market, we mean services marketed and sold on a standardized basis to residential customers, small businesses, and other end-user customers such as schools and libraries. “Schools” would include institutions of higher education to the extent that they purchase these standardized retail services. For purposes of this definition, “mass market” also includes broadband internet access service purchased with the support of the E-rate and Rural Healthcare programs, as well as any broadband internet access service offered using networks supported by the Connect America Fund (CAF), but does not include enterprise service offerings or special access services, which are typically offered to larger organizations through customized or individually negotiated arrangements.

3. The term “broadband internet access service” includes services provided over any technology platform, including but not limited to wire, terrestrial wireless (including fixed and mobile wireless services using licensed or unlicensed spectrum), and satellite. For purposes of our discussion, we divide the various forms of broadband internet access service into the two categories of “fixed” and “mobile.” With these two categories of services—fixed and mobile—we intend to cover the entire universe of internet access services at issue in the Commission's

prior broadband classification decisions, as well as all other broadband internet access services offered over other technology platforms that were not addressed by prior classification orders. We also make clear that our classification finding applies to all providers of broadband internet access service, as we delineate them here, regardless of whether they lease or own the facilities used to provide the service. “Fixed” broadband internet access service refers to a broadband internet access service that serves end users primarily at fixed endpoints using stationary equipment, such as the modem that connects an end user's home router, computer, or other internet access device to the internet. The term encompasses the delivery of fixed broadband over any medium, including various forms of wired broadband services (

e.g.,

cable, DSL, fiber), fixed wireless broadband services (including fixed services using unlicensed spectrum), and fixed satellite broadband services. “Mobile” broadband internet access service refers to a broadband internet access service that serves end users primarily using mobile stations. Mobile broadband internet access includes, among other things, services that use smartphones or mobile-network-enabled tablets as the primary endpoints for connection to the internet. The term also encompasses mobile satellite broadband services. We note that “public safety services” as defined in Section 337(f)(1) would not meet the definition of “broadband internet access service” subject to the rules herein given that “such services are not made commercially available to the public by the provider” as a mass-market retail service.

4. As the Commission found in 2010, broadband internet access service does not include services offering connectivity to one or a small number of internet endpoints for a particular device,

e.g.,

connectivity bundled with e-readers, heart monitors, or energy consumption sensors, to the extent the service relates to the functionality of the device. To the extent these services are provided by ISPs over last-mile capacity shared with broadband internet access service, they would be non-broadband internet access service data services (formerly specialized services). As the Commission found in both 2010 and 2015, non-broadband internet access service data services do not fall under the broadband internet access service category. Such services generally are not used to reach large parts of the internet; are not a generic platform, but rather a specific applications-level service; and use some form of network management to isolate the capacity used by these services from that used by broadband internet access services. Further, we observe that to the extent ISPs “use their broadband infrastructure to provide video and voice services, those services are regulated in their own right.”

5. Broadband internet access service also does not include virtual private network (VPN) services, content delivery networks (CDNs), hosting or data storage services, or internet backbone services (if those services are separate from broadband internet access service), consistent with past Commission precedent. The Commission has historically distinguished these services from “mass market” services, as they do not provide the capability to transmit data to and receive data from all or substantially all internet endpoints. We do not disturb that finding here. Consistent with past Commissions, we note that the transparency rule we adopt today applies only so far as the limits of an ISP's control over the transmission of data to or from its broadband customers.

6. Finally, we observe that to the extent that coffee shops, bookstores, airlines, private end-user networks such as libraries and universities, and other businesses acquire broadband internet access service from an ISP to enable patrons to access the internet from their respective establishments, provision of such service by the premise operator would not itself be considered a broadband internet access service unless it was offered to patrons as a retail mass market service, as we define it here. Although not bound by the transparency rule we adopt today, we encourage premise operators to disclose relevant restrictions on broadband service they make available to their patrons. Likewise, when a user employs, for example, a wireless router or a Wi-Fi hotspot to create a personal Wi-Fi network that is not intentionally offered for the benefit of others, he or she is not offering a broadband internet access service under our definition, because the user is not marketing and selling such service to residential customers, small business, and other end-user customers such as schools and libraries.

2. Broadband Internet Access Service is an Information Service Under the Act

7. In deciding how to classify broadband internet access service, we find that the best reading of the relevant definitional provisions of the Act supports classifying broadband internet access service as an information service. Section 3 of the Act defines an “information service” as “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.” Section 3 defines a “telecommunications service,” by contrast, as “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.” Finally, Section 3 defines “telecommunications”—used in each of the prior two definitions—as “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.” Prior to the

Title II Order

the Commission had long interpreted and applied these terms to classify various forms of internet access service as information services—a conclusion affirmed as reasonable by the Supreme Court in

Brand X.

Our action here simply returns to that prior approach.

8. When interpreting a statute it administers, the Commission, like all agencies, “must operate `within the bounds of reasonable interpretation.' And reasonable statutory interpretation must account for both `the specific context in which . . . language is used' and `the broader context of the statute as a whole.' ” Below, we first explore the meaning of the “capability” contemplated in the statutory definition of “information service,” and find that broadband internet access service provides consumers the “capability” to engage in all of the information processes listed in the information service definition. We also find that broadband internet access service likewise provides information processing functionalities itself, such as DNS and caching, which satisfy the capabilities set forth in the information service definition. We then address what “capabilities” we believe are being “offered” by ISPs, and whether these are reasonably viewed as separate from or inextricably intertwined with transmission, and find that broadband internet access service offerings inextricably intertwine these information processing capabilities with transmission.

9. We find that applying our understanding of the statutory definitions to broadband internet access service as it is offered today most soundly leads to the conclusion that it

is an information service. Although the internet marketplace has continued to develop in the years since the earliest classification decisions, broadband internet access service offerings still involve a number of “capabilities” within the meaning of the Section 3 definition of information services, including critical capabilities that all ISP customers must use for the service to work as it does today. While many popular uses of the internet have shifted over time, the record reveals that broadband internet access service continues to offer information service capabilities that typical users both expect and rely upon. Indeed, the basic nature of internet service—“[p]rovid[ing] consumers with a comprehensive capability for manipulating information using the internet via high-speed telecommunications”—has remained the same since the Supreme Court upheld the Commission's similar classification of cable modem service as an information service twelve years ago.

10. A body of precedent from the courts and the Commission served as the backdrop for the 1996 Act and informed the Commission's original interpretation and implementation of the statutory definitions of “telecommunications,” “telecommunications service,” and “information service.” The classification decisions in the

Title II Order

discounted or ignored much of that precedent. Without viewing ourselves as formally bound by that prior precedent, we find it eminently reasonable, as a legal matter, to give significant weight to that pre-1996 Act precedent in resolving how the statutory definitions apply to broadband internet access service, enabling us to resolve statutory ambiguity in a manner that we believe best reflects Congress's understanding and intent. Our analysis thus is not at odds with the statement in

USTelecom

that the 1996 Act definitions were not “intended to freeze in place the Commission's existing classification of various services.” Consistent with this approach as a traditional tool of statutory interpretation, we reject arguments that suggest that we should disregard this precedent largely out-of-hand. More generally, of course, this precedent—

Brand X

in particular—demonstrates that the Act does not compel a telecommunications service classification.

a. Broadband Internet Access Service Information Processing Capabilities

11. We begin by evaluating the “information service” definition and conclude that it encompasses broadband internet access service. Broadband internet access service includes “capabilit[ies]” meeting the information service definition under a range of reasonable interpretations of that term. In other contexts, the Commission has looked to dictionary definitions and found the term “capability” to be “broad and expansive,” including the concepts of “potential ability” and “the capacity to be used, treated, or developed for a particular purpose.” Because broadband internet access service necessarily has the capacity or potential ability to be used to engage in the activities within the information service definition—“generating, acquiring, storing, transforming, processing, retrieving, utilizing, or making available information via telecommunications”—we conclude that it is best understood to have those “capabilit[ies].” The record reflects that fundamental purposes of broadband internet access service are for its use in “

generating”

and “

making available”

information to others, for example through social media and file sharing; “

acquiring”

and “

retrieving”

information from sources such as websites and online streaming and audio applications, gaming applications, and file sharing applications; “

storing”

information in the cloud and remote servers, and via file sharing applications; “

transforming”

and “

processing”

information such as by manipulating images and documents, online gaming use, and through applications that offer the ability to send and receive email, cloud computing and machine learning capabilities; and “

utilizing”

information by interacting with stored data. These are just a few examples of how broadband internet access service enables customers to generate, acquire, store, transform, process, retrieve, utilize, and make available information. These are not merely incidental uses of broadband internet access service—rather, because it not only has “the capacity to be used” for these “particular purpose[s]” but was designed and intended to do so, we find that broadband internet access is best interpreted as providing customers with the “capability” for such interactions with third party providers.

12. We also find that broadband internet access is an information service irrespective of whether it provides the

entirety

of any end user functionality or whether it provides end user functionality in tandem with edge providers. We do not believe that Congress, in focusing on the “offering of a capability,” intended the classification question to turn on an analysis of which capabilities the end user selects. Further, we are unpersuaded by commenters who assert that in order to be considered an “information service,” an ISP must not only offer customers the “capability” for interacting with information that may be offered by third parties (“click-through”), but must also provide the ultimate content and applications themselves. Although there is no dispute that many edge providers likewise perform functions to facilitate information processing capabilities, they

all

depend on the combination of information-processing and transmission that ISPs make available through broadband internet access service. The fundamental purpose of broadband internet access service is to “enable a constant flow of computer-mediated communications between end-user devices and various servers and routers to facilitate interaction with online content.”

13. From the earliest decisions classifying internet access service, the Commission recognized that even when ISPs enable subscribers to access third party content and services, that can constitute “a capability for generating, acquiring, storing, transforming, processing, retrieving, utilizing, or making available information via telecommunications.” As the Commission explained in the

Stevens Report,

“[s]ubscribers can retrieve files from the World Wide Web, and browse their contents,

because

their service provider offers the `capability for . . . acquiring, . . . retrieving [and] utilizing . . . information.' ” Attempts to distinguish the Commission's classification precedent thus are unfounded insofar as they fail to account for this aspect of the Commission's analysis in those orders. Thus, even where an ISP enables end-users to access the content or applications of a third party, the Commission nonetheless found that constituted the requisite information service “capability.” When the

Title II Order

attempted to evaluate customer perception based on their usage of broadband internet access service, it failed to persuasively grapple with the relevant implications of prior Commission classification precedent. The

Title II Order

argued that broadband internet access service primarily is used to access content, applications, and services from third parties unaffiliated with the ISP in support of the view that customers perceive it as a separate offering of telecommunications. The

Title II Order

offers no explanation as to why its narrower view of “capability”

was more reasonable than the Commission's previous, long-standing view (other than seeking to advance the classification outcome that

Order

was driving towards). Consequently, the

Title II Order

essentially assumed away the legal question of whether end-users perceive broadband internet access service as offering them the “capability for . . . acquiring, . . . retrieving [and] utilizing . . . information” under the broader reading of “capability” in prior Commission precedent.

14. But even if “capability” were understood as requiring more of the information processing to be performed by the classified service itself, we find that broadband internet access service meets that standard. Not only do ISPs offer end users the capability to interact with information online in each and every one of the ways set forth above, they also do so through a variety of functionally integrated information processing components that are part and parcel of the broadband internet access service offering itself. In particular, we conclude that DNS and caching functionalities, as well as certain other information processing capabilities offered by ISPs, are integrated information processing capabilities offered as part of broadband internet access service to consumers today. In addition to DNS and caching, the record reflects that ISPs may also offer a variety of additional features that consist of information processing functionality inextricably intertwined with the underlying service. These additional features include, and are not limited to: email, speed test servers, backup and support services, geolocation-based advertising, data storage, parental controls, unique programming content, spam protection, pop-up blockers, instant messaging services, on-the-go access to Wi-Fi hotspots, and various widgets, toolbars, and applications. While we do not find the offering of these information processing capabilities determinative of the classification of broadband internet access service, their inclusion in the broadband internet access service, and the capabilities and functionalities necessary to make these features possible, further support the “information service” classification.

15.

DNS.

We find that DNS is an indispensable functionality of broadband internet access service. While we accept that DNS is not necessary for transmission, we reject assertions that it is not indispensable to the broadband internet access service customers use—and expect—today. DNS is a core function of broadband internet access service that involves the capabilities of generating, acquiring, storing, transforming, processing, retrieving, utilizing and making available information. DNS is used to facilitate the information retrieval capabilities that are inherent in internet access. DNS allows “`click through' access from one web page to another, and its computer processing functions analyze user queries to determine which website (and server) would respond best to the user's request.” And “[b]ecause it translates human language (

e.g.,

the name of a website) into the numerical data (

i.e.,

an IP address) that computers can process, it is indispensable to ordinary users as they navigate the internet.” Without DNS, a consumer would not be able to access a website by typing its advertised name (

e.g.,

fcc.gov

or

cnn.com

). The

Brand X

Court recognized the importance of DNS, concluding that “[f]or an internet user, `DNS is a must. . . . [N]early all of the internet's network services use DNS. That includes the World Wide Web, electronic mail, remote terminal access, and file transfer.' ” While ISPs are not the sole providers of DNS services, the vast majority of ordinary consumers rely upon the DNS functionality provided by their ISP, and the absence of ISP-provided DNS would fundamentally change the online experience for the consumer. We also observe that DNS, as it is used today, provides more than a functionally integrated address-translation capability, but also enables other capabilities critical to providing a functional broadband internet access service to the consumer, including for example, a variety of underlying network functionality information associated with name service, alternative routing mechanisms, and information distribution.

16. The treatment of similar functions in MFJ precedent bolsters our conclusion. Despite the fact that the telecommunications management exception (and information service definition more broadly) was drawn most directly from the MFJ, the

Title II Order

essentially ignored MFJ precedent when concluding that DNS fell within the statutory telecommunications management exception. In addition, even the

Title II Order'

s limited use of

Computer Inquiries

precedent focused mostly on relatively high-level Commission statements about the general sorts of capabilities that could be basic (or adjunct-to-basic) or drew analogies to specific holdings that are at best ambiguous as to their application to broadband internet access service. When analyzing “gateway” functionalities by which BOCs would provide end-users with access to third party information services, the MFJ court found that “address translation,” which enabled “the consumer [to] use an abbreviated code or signal . . . in order to access the information service provider” such as through “the translation of a mnemonic code into [a] telephone number,” rendered gateways an information service. We recognize that gateway functionalities and broadband internet access service are not precisely coextensive in scope. We do, however, find similarities between functionalities such as address translation and storage and retrieval to key functionalities provided by ISPs as part of broadband internet access service, and we conclude the court found such gateway and similar functionalities independently sufficient to warrant an information service classification under the MFJ. The “address translation” gateway function appears highly analogous to the DNS function of broadband internet access service, which enables end users to use easier-to-remember domain names to initiate access to the associated IP addresses of edge providers. That MFJ precedent, neglected by the

Title II Order,

thus supports our finding that the inclusion of DNS in broadband internet access service offerings likewise renders that service an information service. We rely on this analogy between DNS and particular functions classified under pre-1996 Act precedent not because the technologies are identical in all particulars, but because they share the same relevant characteristics for purposes of making a classification decision under the Act. Given the close fit between DNS and the address translation function classified as an information service under the MFJ coupled with the fact that the statutory information service definition (and telecommunications management exception) was drawn more directly from the MFJ, we find the MFJ precedent entitled to more weight than analogies to

Computer Inquiries

precedent. We thus are not persuaded by arguments seeking to analogize DNS to directory assistance, which the Commission classified as “adjunct-to-basic” under the

Computer Inquiries.

17. We thus find that the

Title II Order

erred in finding that DNS functionalities fell within the telecommunications systems management exception to the definition of “information service.” That exception from the statutory information service definition was drawn from the language of the MFJ, and was understood as “directed at internal operations, not at services for

customers or end users.” The court's definition of information services excluded capabilities “for the management, control, or operation of a telecommunication system or the management of a telecommunications service.” Under the Communications Act, the definition of “information services” includes an identically-worded “telecommunications management” exception. Commission precedent and legislative history likewise recognize that the definition was drawn from the MFJ. We interpret the concepts of “management, control, or operation” in the telecommunications management exception consistent with that understanding. Applying that interpretation, we find the record reflects that little or nothing in the DNS look-up process is designed to help an ISP “manage” its network; instead, DNS functionalities “provide stored information to

end

users to help them navigate the internet.” As AT&T explains: “When an end user types a domain name into his or her browser and sends a DNS query to an ISP, . . . the ISP . . . converts the human-language domain name into a numerical IP address, and it then conveys that information back to the end user . . . [who] (via his or her browser) thereafter sends a follow-up request for the internet resources located at that numerical IP address.” DNS does not merely “manage” a telecommunications service, as some commenters assert, but rather is a function that is useful and essential to providing internet access for the ordinary consumer. We are persuaded that “[w]ere DNS simply a management function, this would not be the case.” Comparing functions that

would

fall within the exception illustrates the distinction. For example, in contrast to DNS's interaction with users and their applications, “non-user, management-only protocols might include things such as Simple Network Management Protocol (SNMP), Network Control Protocol (NETCONF), or DOCSIS bootfiles for controlling the configuration of cable modems.” These protocols support services that manage the network independent of the transmission of information initiated by a user. Other functions that would fall into the telecommunications systems management exception might include information systems for account management and billing, configuration management, and the monitoring of failures and other state information, and to keep track of which addresses are reachable through each of the interconnected neighboring networks.

18. The

Title II Order

drew erroneous conclusions from

Computer Inquiries

precedent and too quickly rejected objections to its treatment of DNS as meeting the telecommunications management exception. The same shortcomings are present in the

Title II Order'

s analysis of caching, as well. Under the

Computer Inquiries

framework, the Commission held that some capabilities “may properly be associated with basic [common carrier] service without changing its nature, or with an enhanced service without changing the classification of the latter as unregulated under Title II of the Act.” These commonly came to be known as “adjunct” capabilities. The Commission has held that functions it had classified as “adjunct-to-basic” under the

Computer Inquiries

framework will fall within the statutory telecommunications management exception to the information service definition. Drawing loose analogies to certain functions described as adjunct-to-basic under Commission precedent, the

Title II Order

held that DNS fell within the telecommunications management exception.

19. The

Title II Order

incorrectly assumed that so long as a functionality was, in part, used in a manner that could be viewed as adjunct-to-basic, it necessarily was adjunct-to-basic regardless of what the functionality otherwise accomplished. In addition to the MFJ precedent, Bureau precedent similarly has observed that adjunct-to-basic capabilities do not include functions “useful to end users, rather than carriers.” Given the lack of ambiguity in the MFJ's holding in this regard, we find it more reasonable to interpret this precedent to call for a similar requirement that “adjunct to basic” services do not include services primarily useful to end-users, and reject arguments to the contrary. Although confronted with claims that DNS is, in significant part, designed to be useful to end-users rather than providers, the

Title II Order

nonetheless decided that it fell within the telecommunications management exception. The same is true of the

Title II Order'

s treatment of caching. While conceding that DNS, as well as other functions like caching, “do provide a benefit to subscribers,” the

Title II Order

held that they nonetheless fell within the telecommunications management exception because it found some aspect of their operation also was of use to providers in managing their networks. This expansive view of the telecommunications management exception—and associated narrowing of the scope of information services—is a transposition of the analytical approach embodied in the MFJ and

Computer Inquiries;

under the approach in the pre-1996 Act precedent, the analysis would instead begin with the broad language of the information service or enhanced service definitions, generally excluding particular functions only if the purpose served clearly was narrowly focused on facilitating bare transmission. The Commission and the courts made clear the narrow scope of the `adjunct-to-basic' or `telecommunications management' categories in numerous decisions in many different contexts.). Notably, the focus remains on the purpose or use of the specific function in question and not merely whether the resulting service, as a whole, is useful to end-users.

20. The

Title II Order

also put misplaced reliance on

Computer Inquiries

adjunct-to-basic precedent from the traditional telephone service context as a comparison when evaluating broadband internet access service functionalities. Because broadband internet access service was not directly addressed in pre-1996 Act

Computer Inquiries

and MFJ precedent, analogies to functions that were classified under that precedent must account for potentially distinguishing characteristics not only in terms of technical details but also in terms of the regulatory backdrop. The 1996 Act enunciates a policy for the internet that distinguishes broadband internet access from legacy services like traditional telephone service. The 1996 Act explains that it is federal policy “to preserve the vibrant and competitive free market that presently exists for the internet and other interactive computer services, unfettered by Federal or State regulation.” The application of potentially ambiguous precedent to broadband internet access service should be informed by how well—or how poorly—it advances that deregulatory statutory policy. We find that our approach to that precedent, which results in an information service classification of broadband internet access service, better advances that deregulatory policy than the approach in the

Title II Order,

which led to the imposition of utility-style regulation under Title II.

21. The regulatory history of traditional telephone service also informs our understanding of

Computer Inquiries

precedent, further distinguishing it from broadband internet access service. Given the long history of common carriage offering of that service by the time of the

Computer Inquiries,

it is understandable that some precedent started with a presumption

that the underlying service was a “basic service.” But similar assumptions would not be warranted in the case of services other than traditional telephone service for which there was no similar longstanding history of common carriage. Thus, not only did the

Title II Order

rely on specific holdings that are at best ambiguous in their analogy to technical characteristics of broadband internet access service, but it failed to adequately appreciate key regulatory distinctions between traditional telephone service and broadband internet access service. Thus, for example, the fact that the adjunct-to-basic classification of directory assistance arose in the traditional telephone context likewise persuades us to give it relatively little weight here as an analogy to DNS, and we reject arguments to the contrary.

22.

Caching.

We also conclude that caching, a functionally integrated information processing component of broadband internet access service, provides the capability to perform functions that fall within the information service definition. As the record reflects, “[c]aching does much more than simply enable the user to obtain more rapid retrieval of information through the network; caching depends on complex algorithms to determine what information to store where and in what format.” This requires “extensive information processing, storing, retrieving, and transforming for much of the most popular content on the internet,” and as such, caching involves storing and retrieving capabilities required by the “information service” definition. The Court affirmed this view in

Brand X,

finding “reasonable” the “Commission's understanding” that internet service “facilitates access to third-party web pages by offering consumers the ability to store, or `cache,' popular content on local computer servers,” which constitutes “the `capability for . . . acquiring, [storing] . . . retrieving [and] utilizing information.' ”

23. We find that ISP-provided caching does not merely “manage” an ISP's broadband internet access service and underlying network, it enables and enhances consumers' access to and use of information online. The record shows that caching can be realized as part of a service, such as DNS, which is predominantly to the benefit of the user (DNS caching). We disagree with assertions in the record that suggest that ISP-provided caching is not a vital part of broadband internet access service offerings, as it may be stymied by the use of HTTPS encryption. Caching can also be realized in terms of content that can be accumulated by the ISP through non-confidential (

i.e.,

non-encrypted) retrieval of information from websites (Web caching). In this case, the user benefits from a rapid retrieval of information from a local cache or repository of information while the ISP benefits from less bandwidth resources used in the retrieval of data from one or more destinations. DNS and Web caching are functions provided as part and parcel of the broadband internet access service. When ISPs cache content from across the internet, they are not performing functions, like switching, that are instrumental to pure transmission, but instead storing third party content they select in servers in their own networks to enhance access to information. The record reflects that without caching, broadband internet access service would be a significantly inferior experience for the consumer, particularly for customers in remote areas, requiring additional time and network capacity for retrieval of information from the internet. Thus, because caching is useful to the consumer, we conclude that the

Title II Order

erred in incorrectly categorizing caching as falling within the telecommunications system management exception to the definition of “information service.”

24. In addition, the

Title II Order'

s failure to consider applicable MFJ precedent led to mistaken analogies when it concluded that caching fell within the statutory telecommunications management exception. In relevant precedent, the MFJ court observed that the information service restriction generally “prohibits the [BOCs] from `storing' and `retrieving' information,” but identified “quite distinct settings in which storage capabilities of the [BOCs] could be used in the information services market.” One of the categories of storage and retrieval identified by the court appears highly comparable to caching. That category involved BOC provision of “storage space in their gateways for databases created by others” such as “information service providers and end users,” making “communication more efficient by moving information closer to the end user, thereby reducing transmission costs.” This functionality—recognized as an information service by the MFJ court—appears highly analogous to caching, and lends historical support to our view that the caching functionality within broadband internet access service is best understood as rendering broadband internet access service an information service. The first category the court identified was “very short term storage,” including, among other things, “the basic packet switching function,” which “involves the breakdown of data or voice communications into small bits of information that are then collected and transmitted between nodes,” involving “constant storage, error checking, and retransmission, as required for accurate transmission.” Although the court was not entirely clear, it seemed to suggest that such functions were not information services under the MFJ. This category appears to bear little similarity to caching, however. The third category of “storage and retrieval” information service functions identified by the court would include the BOC's provision of “voice messaging, voice storage and retrieval, and electronic mail.” Because that category does not appear as analogous to caching as the category identified by the court and described above, nor was it relied upon in the

Title II Order'

s discussion of caching, we do not focus on that third category in our discussion here.

25. Ignoring that MFJ precedent, the

Title II Order

erred in seeking to analogize caching to “ `store and forward technology [used] in routing messages through the network as part of a basic service' ” mentioned in the

Computer II Final Decision.

In fact, consistent with the MFJ court's identification of distinct uses of storage and forwarding, the cited portion of the

Computer II Final Decision

recognized that “the kind of enhanced store and forward services that can be offered are many and varied.” In that regard, the

Computer II Final Decision

distinguished “[t]he offering of store and forward

services”

from “store and forward

technology,”

explaining that “[m]essage or packet switching, for example, is a store and forward technology that may be employed in providing basic service.” Reading that discussion in full context and in harmony with subsequent MFJ precedent, the reference in the

Computer II Final Decision

to “store and forward technology” appears better understood as mirroring a category of storage and retrieval of information that the MFJ court suggested was not an information service—in particular, “the basic packet switching function, . . . [which] involves the breakdown of data or voice communications into small bits of information that are then collected and transmitted between nodes.” That category of activity relied upon in the

Title II Order

thus actually appears to be barely or not at all analogous to caching. We instead find more persuasive the

MFJ court's information service treatment of BOC provision of “storage space in their gateways for databases created by others” such as “information service providers and end users”—a distinct category of storage and retrieval functionality that is a close fit to caching. We are unpersuaded by claims that this MFJ precedent only is analogous to CDNs and not “transparent caching” based on asserted differences in how it is determined what content will be stored in each scenario. Although the factual scenario discussed in the MFJ anticipated end-users or information service providers electing what information to store, and that fact may have partially informed the court's decision whether to ultimately allow BOCs to provide that capability notwithstanding its classification as an information service, we do not read the underlying classification as turning on that issue. Further, in addition to the distinctions between caching and store-and-forward technology acknowledged even in this filing, Peha Dec. 7, 2017

Ex Parte

Letter at 4, we find additional shortcomings in how the

Title II Order

relied on adjunct-to-basic precedent.

b. ISPs' Service Offerings Inextricably Intertwine Information Processing Capabilities With Transmission

26. Having established that broadband internet access service has the information processing capabilities outlined in the definition of “information service,” the relevant inquiry is whether ISPs' broadband internet access service offerings make available information processing technology inextricably intertwined with transmission. Below we examine both how consumers perceive the offer of broadband internet access service, as well as the nature of the service actually offered by ISPs, and conclude that ISPs are best understood as offering a service that inextricably intertwines the information processing capabilities described above and transmission.

27. We begin by considering the ordinary customer's perception of the ISP's offer of broadband internet access service. As

Brand X

explained, “[i]t is common usage to describe what a company `offers' to a consumer as what the consumer perceives to be the integrated finished product.” ISPs generally market and provide information processing capabilities and transmission capability together as a single service. Therefore, it is not surprising that consumers perceive the offer of broadband internet access service to include more than mere transmission, and that customers want and pay for functionalities that go beyond mere transmission. As Cox explains, “[w]hile consumers also place significant weight on obtaining a reliable and fast internet connection, they view those attributes as a means of enabling these capabilities to interact with information online, not as ends in and of themselves.” Indeed, record evidence confirms that consumers highly value the capabilities their ISPs offer to acquire information from websites, utilize information on the internet, retrieve such information, and otherwise process such information. NHMC's argument, based on what it asserts to be a representative sample of consumer complaints filed with the Commission, is not persuasive. NHMC's methodology relied on Natural Language Processing (NLP) to determine words that co-occur in such complaints, and then used “iterative clustering algorithms” to “ma[p] connections among them.” Neither NHMC's methodology nor the representative extracts of the complaints NHMC submitted demonstrate that individual complaints about particular aspects of service reflect how a customer would perceive service offerings as a whole. Indeed, the sample of complaints attached by NHMC features a broad set of issues, ranging widely from questions about speed to “losing my internet connection,” “charg[ing] extra for your services,” “interrupt[ing] the service,” “bully[ing] me into share plans,” “Google arbitrarily engag[ing] in monopolistic practices,” “charg[ing] me modem rental fee,” or “basically no technical support.” We further note that to the extent that perceived speed is a common complaint, that does not mean consumers view broadband internet access service as a pure transmission service. A consumer's perceived speed for many activities (such as web browsing) depends on information-processing elements of the service like DNS and caching; indeed, caching's primary consumer benefit is allowing a more rapid retrieval of information from a local cache (increasing the perceived speed of a consumer's connection). Moreover, the Commission has never relied on such complaints to identify what a service is. And for good reason: We expect consumer complaints about problems with a service—not every aspect of it. Indeed, applying such a methodology would lead to absurd results: Should we redefine the public switched network based on the millions of robocall complaints we get each year or the rural-call-completion problems that we know are too prevalent? Of course not.

28. This view also accords with the Commission's historical understanding that “[e]nd users subscribing to . . . broadband internet access service expect to receive (and pay for) a finished, functionally integrated service that provides access to the internet. End users do not expect to receive (or pay for) two distinct services—both internet access service and a distinct transmission service, for example.” While the

Title II Order

dwells at length on the prominence of transmission speed in ISP marketing, it makes no effort to compare that emphasis to historical practice. In fact, ISPs have been highlighting transmission speed in their marketing materials since long before the

Title II Order.

The very first report on advanced telecommunication capability pursuant to Section 706(b) of the 1996 Act, released in 1999, cited ISPs' marketing of their internet access service speed. ISPs' inclusion of speed information in their marketing also was acknowledged by the Court in

Brand X,

which nonetheless upheld the Commission's information service classification as reasonable. Indeed, consideration of ISP marketing practices has been part of the backdrop of all of the Commission's decisions classifying broadband internet access service as an information service and thus cannot justify a departure from the historical classification of broadband internet access service as an information service.

29. The

Title II Order'

s reliance on ISP marketing also assumes that it provides a complete picture of what consumers perceive as the finished product. First, the record reflects that ISP marketing of broadband encompasses features beyond speed and reliability. Further, because all broadband internet access services rely on DNS and commonly also rely on caching by ISPs, to the extent that those capabilities, in themselves, do not provide a point of differentiation among services or providers, it would be unsurprising that ISPs did not feature them prominently in their marketing or advertising, particularly to audiences already familiar with broadband internet access service generally. Indeed, speed and reliability are not exclusive to telecommunications services; rather, the record reflects that speed and reliability are crucial attributes of an information service. As such, we reject assertions that speed and reliability are only characteristics of telecommunications services and further note that ISPs market these aspects because they can be differentiated, unlike DNS or caching. Consequently, the mere fact that broadband internet access service

marketing often focuses on characteristics, such as transmission speed, by which services and providers can be differentiated sheds little to no light on whether consumers perceive broadband internet access service as inextricably intertwining that data transmission with information service capabilities. Neither the discussion of the consumer's perspective by Justice Scalia nor that in the

Title II Order

identifies good reasons to depart from the Commission's prior understanding that broadband internet access is a single, integrated information service. Justice Scalia contended that how customers perceive cable modem service is best understood by considering the services for which it would be a substitute—in his view at the time, dial-up internet access and digital subscriber line (DSL) service over telephone networks. However, dial-up internet access has substantially diminished in marketplace significance in the subsequent years. In addition, the legal compulsion for facilities-based carriers to offer broadband transmission on a common carrier basis was eliminated in 2005. Fixed and mobile wireless broadband internet access service have grown to play a much more prominent role in the broadband internet access service marketplace, along with satellite broadband internet access service, none of which ever was under a legal compulsion to offer broadband transmission on a common carrier basis—nor, prior to the

Title II Order,

were they interpreted as voluntarily doing so. Consequently, whatever might have been arguable at the time of

Brand X,

the service offerings in the marketplace as it developed thereafter provide no reason to expect that consumers “inevitabl[y]” would view broadband internet access service as involving “

both

computing functionality

and

the physical pipe” as separate offerings based on comparisons to the likely alternatives.

30. Separate and distinct from our finding that an ISP “offers” an information service from the consumer's perspective, we find that as a factual matter, ISPs offer a single, inextricably intertwined information service. The record reflects that information processes must be combined with transmission in order for broadband internet access service to work, and it is the combined information processing capabilities and transmission functions that an ISP offers with broadband internet access service. Thus, even assuming that any individual consumer could perceive an ISP's offer of broadband internet access service as akin to a bare transmission service, the information processing capabilities that are actually offered as an integral part of the service make broadband internet access service an information service as defined by the Act. As such, we reject commenters' assertions that the primary function of ISPs is to simply transfer packets and not process information.

31. The inquiry called for by the relevant classification precedent focuses on the nature of the service offering the provider makes, rather than being limited to the functions within that offering that particular subscribers do, in fact, use or that third parties also provide. As the Commission recognized in the

Cable Modem Order,

internet access service was appropriately classified as an offering of the capabilities with the definition of an information service “regardless of whether subscribers use all of the functions provided as part of the service.” The

Title II Order

erroneously contended that, because functions like DNS and caching potentially could be provided by entities other than the ISP itself, those functions should not be understood as part of a single, integrated information service offered by ISPs. However, the fact that some consumers obtain these functionalities from third-party alternatives is not a basis for ignoring the capabilities that a broadband provider actually “offers.” The

Title II Order

gave no meaningful explanation why a contrary, narrower interpretation of “offer” was warranted other than, implicitly, its seemingly end-results driven effort to justify a telecommunications service classification of broadband internet access service.

32. Our findings today are consistent with classification precedent prior to the

Title II Order,

which consistently found that ISPs offer a single, integrated service. Although we find the pre-1996 Act

classification

precedent relevant to our classification of broadband internet access service, we reject the view that Congress would have expected classification under the 1996 Act's statutory definitions to be tied to the substantive common carrier transmission

requirements

imposed under those frameworks. We conclude that the best view of the text and structure of the Act undercuts arguments that Congress sought to preserve the substance of pre-1996 Act regulations through the definitions it adopted. Instead, where Congress sought to address substantive requirements akin to those in the MFJ and

Computer Inquiries,

it did so by adopting subjective obligations in the 1996 Act—even if not identical to the pre-1996 Act requirements—and subject to their own Congressionally specified standards for when and to what entities they apply. In addition, the wholesale service focus of substantive MFJ and

Computer Inquiries

common carrier transmission obligations also distinguishes them from the retail service we classify here, likewise undermining any claimed relevance of those pre-1996 Act transmission requirements to our classification decision. The Commission recognized, for example, that the transmission underlying broadband internet access required by the

Computer Inquiries

to be offered on an unbundled, common carrier basis and provided to ISPs was not a “retail” service within the meaning of Section 251(c)(4) resale requirements. Nor did such a common carrier transmission service itself enable access to the internet, even if purchased by end-users. By comparison, under the

Computer Inquiries,

the finished service offered to end-users relying on the required common carrier transmission as an input was regulated as an enhanced service, not a common carrier offering, even when offered by the facilities-based carrier's subsidiary. Given our focus here on the finished retail broadband internet access service, we see little relevance to prior regulatory requirements that were imposed to ensure competing providers had access to a wholesale input in the form of a compelled common carriage offering of bare transmission that did not itself provide internet access. Even the early classification analysis in the

Stevens Report

recognized that “[i]n offering service to end users” ISPs “do more than resell [ ] data transport services. They conjoin the data transport with data processing, information provision, and other computer-mediated offerings, thereby creating an information service.” In

Brand X,

the Court rejected claims that “[w]hen a consumer . . . accesses content provided by parties other than the cable company” that “consumer uses `pure transmission.' ” Subsequent Commission decisions involving other forms of broadband internet access likewise all concluded that the broadband internet access service was a single, integrated service that did not involve a stand-alone offering of telecommunications. Although parties have, over time, held various views regarding the proper classification of broadband internet access services, the mere fact that a party held such a view in the past, or holds such a view today, does not render a Commission decision confirming a particular view “moot,”

since a private party's subjective view is not authoritative. The Court further found that “the high-speed transmission used to provide cable modem service is a functionally integrated component of that service because it transmits data only in connection with the further processing of information and is necessary to provide internet service.” This distinction makes broadband internet access service fundamentally different than standard telephone service, which the Supreme Court noted does not become an “information service” merely because its transmission service may be “trivially affected” by some additional capability such as voicemail. Where the addition of some further capability has appeared to have only a trivial effect on the nature of a service, the Commission has previously declined requests for reclassification. Due to the functionally integrated nature of broadband internet access service, however, we reject claims that those decisions call for a different approach than we adopt here. Likewise, the outcome in the Bureau-level

Cisco WebEx Order

accords with our approach, given the finding that the information service capabilities more than trivially affected the transmission capability in the scenario addressed there. Contrary to some arguments, the Bureau had no need to—and did not—address the classification of other service scenarios, and we reject arguments for a different classification approach that are premised on assumptions about how those unaddressed scenarios would have been analyzed or classified. The core, essential elements of these prior analyses of the functional nature of internet access remain persuasive as to broadband internet access service today. We adhere to that view notwithstanding arguments that some subset of the array of internet access uses identified in the

Stevens Report

or subsequent decisions either are no longer as commonly used, or occur more frequently today. Even at the time of the

Cable Modem Order

the Commission recognized the role of user-generated content, and its decision in no way hinged on distinctions in how retail customers of cable modem service used that service in that respect.

33. We disagree with commenters who assert that ISPs necessarily offer both an information service and a telecommunications service because broadband internet access service includes a transmission component. In providing broadband internet access service, an ISP

makes use

of telecommunications—

i.e.,

it provides information-processing capabilities “via telecommunications”—but does not separately

offer

telecommunications on a stand-alone basis to the public. By definition,

all

information services accomplish their functions “via telecommunications,” and as such, broadband internet access service has always had a telecommunications component intrinsically intertwined with the computer processing, information provision, and computer interactivity capabilities an information service offers. We observe that placing information in IP packets does not change the form of information. We find that the transmission of IP packets is transmission of the user's choosing, and also agree that “[c]hanging the packet structure of an IP packet from IPv4 to IPv6” does not change the form of the information. As just one example, in support of its classification decision, the

Title II Order

notes that it is technically possible for a transmission component underlying broadband internet access service to be separated out and offered on a common carrier basis. The same would be equally true of many information services, however, given that the information service capabilities are, by definition, available “via telecommunications.” Indeed, service providers, who are in the best position to understand the inputs used in broadband internet access service, do not appear to dispute that the “via telecommunications” criteria is satisfied even if also arguing that they are

not providing

telecommunications to end-users. For example, ISPs typically transmit traffic between aggregation points on their network and the ISPs' connections with other networks. Whether self-provided by the ISP or purchased from a third party, that readily appears to be transmission between or among points selected by the ISP of traffic that the ISP has chosen to have carried by that transmission link. We reject as overbroad the claim that “a transmission is `telecommunications' within the meaning of 47 U.S.C. 153(30)

only

if the transmission is capable of communicating with

all

circuit switched devices on the PSTN or has the purpose of facilitating the use of the PSTN without altering its fundamental character as a telephone network.” This claim appears premised on incorporating Section 332's definition of a commercial mobile service (which must be “interconnected” with the “public switched network”) into Section 3 of the Act and drawing from pre-1996 Act precedent using an end-to-end analysis to determine the regulatory jurisdiction of communications traffic to inform the interpretation of the term “points.” But we find no evidence in the text of the statute that Congress intended to import the commercial mobile service definition from one section into another, and our precedent similarly does not countenance such an importation. Nor is the end-to-end analysis the only pre-1996 Act precedent from which the concept of “points” in the “telecommunications” definition might have been drawn so as to unambiguously foreclose our conclusion that “via telecommunications” is satisfied here. Such inclusion of a transmission component does not render broadband internet access services telecommunications services; if it did, the entire category of information services would be narrowed drastically. Because we find it more reasonable to conclude that at least some telecommunications is being used as an input into broadband internet access service—thereby satisfying the “via telecommunications” criteria—we need not further address the scope of the “telecommunications” definition in order to justify our classification of broadband internet access service as an information service. We thus do not comprehensively address other criticisms of the

Title II Order'

s interpretation and applications of the “telecommunications” definition, which potentially could have implications beyond the scope of issues we are considering in this proceeding.

34. The approach we adopt today best implements the Commission's long-standing view that Congress intended the definitions of “telecommunications service” and “information service” to be mutually exclusive ways to classify a given service. As the

Brand X

Court found, the term “offering” in the telecommunications service definition “can reasonably be read to mean a `stand-alone' offering of telecommunications.” Where, as in the case of broadband internet access services, a service involving transmission inextricably intertwines that transmission with information service capabilities—in the form of an integrated information service—there cannot be “a `stand-alone' offering of telecommunications” as required under that interpretation of the telecommunications service definition. This conclusion is true even if the information service could be said to involve the provision of telecommunications as a component of the service. The Commission's historical approach to internet access services carefully navigated that issue, while the

Title II Order,

by contrast, threatened to

usher in a much more sweeping scope of “telecommunications services.”

35. The

Title II Order

interpretation stands in stark contrast to the Commission's historical classification precedent and the views of all Justices in

Brand X.

Beginning with the earliest classification decisions, the Commission found that transmission provided by ISPs outside the last mile was part of an integrated information service. The DSL transmission service previously required to be unbundled by the

Computer Inquiries

rules likewise was limited to the “last mile” connection between the end-user and the ISP. Nor did any Justice in

Brand X

contest the view that, beyond the last mile, cable operators were offering an information service. Indeed, the

Title II Order'

s broad interpretation of “telecommunications service” stands in contrast to the views of Justice Scalia himself, on which the

Title II Order

purports to rely. Justice Scalia was skeptical that a telecommunications service classification of cable modem service would lead to the classification of ISPs as telecommunications carriers based on the transmission underlying their “connect[ions] to other parts of the internet, including internet backbone providers.” Yet the

Title II Order

reached essentially that outcome. The

Title II Order'

s interpretation of the statutory definitions did not merely lead it to classify “last mile” transmission as a telecommunications service. Rather, under the view of the

Title II Order,

even the transmissions underlying an ISP's connections to other parts of the internet, including internet backbone providers, were part of the classified telecommunications service. Even if the

Title II Order'

s classification approach does not technically render the category of information services a nullity, the fact that its view of telecommunications services sweeps so much more broadly than previously considered possible provides significant support for our reading of the statute and the classification decision we make today. That the Commission previously identified policy concerns about internet traffic exchange says nothing about classification, and thus is not to the contrary. Nor did the

Advanced Services

proceedings identify interconnection obligations on providers of xDSL transmission as services necessary to ensure the provision of internet access. Instead, any interconnection obligations identified there were limited to interconnection between providers of common carrier xDSL transmission service and other telecommunications carriers (rather than providers of edge services or non-common carrier backbone services). The cited portion of the

Advanced Services Remand Order

does not even have anything to do with interconnection requirements or the scope of functions in an xDSL-based advanced service. Rather, it analyzed the jurisdiction of the traffic being carried over the service, which, under the traditional end-to-end analysis, was not limited in scope to any given service within a broader communications pathway.

36. In contrast, our approach leaves ample room for a meaningful range of “telecommunications services.” Historically, the Commission has distinguished service offerings that “always and necessarily combine” functions such as “computer processing, information provision, and computer interactivity with data transport, enabling end users to run a variety of applications such as email, and access web pages and newsgroups,” on the one hand, from services “that carriers and end users typically use [ ] for basic transmission purposes” on the other hand. Our interpretation thus stops far short of the view that “

every transmission

of information becomes an information service.” Thus, an offering like broadband internet access service that “always and necessarily” includes integrated transmission and information service capabilities would be an information service. The distinction between services that “always and necessarily” include integrated transmission and information service capabilities and those that do not also highlights a critical difference between internet access service and the service addressed in precedent such as the

Advanced Services Order.

The transmission underlying internet access service that, prior to the

Wireline Broadband Classification Order,

carriers had been required by the

Computer Inquiries

to unbundle and offer as a bare transmission service on a common carrier basis to ensure its availability to competing enhanced service providers—and which did not itself provide internet access—is another specific example of a service that does not “always and necessarily” include integrated transmission and information service capabilities. The Commission naturally recognized at the time that the compelled common carriage offering of bare transmission was a telecommunications service, and we reject the view that such an acknowledgment is inconsistent with, or undercuts our reliance on, precedent classifying internet access service as an integrated information service. In addition, the discussion of xDSL advanced services in the

Advanced Services Order

cited by commenters addressed the transmission service generally. It did not purport to be focused specifically on the use of xDSL transmission in connection with internet access service, rather than addressing the classification of the stand-alone transmission service as a general matter. The Commission's historical interpretation thus gives full meaning to both “information service” and “telecommunications service” categories in the Act.

37. We reject assertions that the analysis we adopt today would necessarily mean that standard telephone service is likewise an information service. The record reflects that broadband internet access service is categorically different from standard telephone service in that it is “

designed with

advanced features, protocols, and security measures so that it can integrate directly into electronic computer systems and enable users to electronically create, retrieve, modify and otherwise manipulate information stored on servers around the world.” Further, “[t]he dynamic network functionality enabling the internet connectivity provided by [broadband internet access services] is fundamentally different from the largely static one dimensional, transmission oriented Time Division Multiplexing (TDM) voice network.” This finding is consistent with past distinctions. Under pre-1996 Act MFJ precedent, for example, although the provision of time and weather services was an information service, when a BOC's traditional telephone service was used to call a third party time and weather service “the Operating Company does not `provide information services' within the meaning of section II(D) of the decree; it merely transmits a call under the tariff.” In other words, the fundamental nature of traditional telephone service, and the commonly-understood purpose for which traditional telephone service is designed and offered, is to provide basic transmission—a fact not changed by its incidental use, on occasion, to access information services. By contrast, the fundamental nature of broadband internet access service, and the commonly-understood purpose for which broadband internet access service is designed and offered, is to enable customers to generate, acquire, store, transform, process, retrieve, utilize, and make available information. In addition, broadband internet access service

includes DNS and caching functionalities, as well as certain other information processing capabilities. As such, we reject assertions that, under the approach we adopt today, any telephone service would be an information service because voice customers can get access to either automated information services or a live person who can provide information.

38. Additionally, efforts to treat the

Stevens Report

as an outlier that should not have been followed in subsequent classification decisions—and should not be followed here—are ultimately unpersuasive. The clear recognition in the

Stevens Report

that the ISPs at issue were themselves providing data transmission as part of their offerings undercuts arguments seeking to distinguish the

Stevens Report

based on the theory that the transmission used to connect to ISPs typically involved common carrier services either directly (via a call to a dial-up ISP using traditional telephone service) or indirectly (with the ISP using common carrier broadband transmission as a wholesale input into its retail information service). While the extent of data transmission provided by the ISPs that were found to be offering information services in the

Stevens Report

might be incrementally less than the transmission provided by the ISPs dealt with in subsequent information service classification decisions, that appears to be at most a difference in degree, rather than a difference in kind, and the record does not demonstrate otherwise. Nor can the

Stevens Report'

s analysis and information service classification be distinguished on the grounds that the ISPs there generally did not own the facilities they used. Although the

Stevens Report

observed that the analysis of whether a single integrated service was being offered was “more complicated when it comes to offerings by facilities-based providers,” it did not prejudge the resolution of that question. Thus, there is no reason to simply assume that it was inappropriate for the Commission to build upon the

Stevens Report

precedent when analyzing service offerings from facilities-based providers beginning in the

Cable Modem Order.

Nor do commenters identify material technical differences when facilities ownership is involved that would mandate a different classification analysis. While the

Stevens Report

recognized that under

Computer Inquires

precedent “offerings by non-facilities-based providers combining communications and computing components should always be deemed enhanced,” had its analysis simply been carrying forward that approach most of its analysis would have been unnecessary (since internet access clearly did combine communications and computing components). Thus, whether or not the more extensive analysis set forth in the

Stevens Report

was necessary to find internet access provided by non-facilities-based ISPs to be an information service, that analysis cannot be said to be a mere relic of the

Computer Inquiries

approach to non-facilities based providers. Finally, our reliance on classification precedent does not rest on the

Stevens Report

alone, but draws from the full range of classification precedent, both pre- and post-1996 Act. This reliance notably includes not only the Commission's classification decisions, but the Supreme Court's subsequent analysis in

Brand X.

And although some commenters criticize the lack of express consideration of the possible application of the telecommunications management exception in the

Stevens Report,

our evaluation of the pre-1996 Act MFJ and

Computer Inquiries

precedent better accords with outcome of that

Report

and the subsequent classification decisions than it does with the

Title II Order

in that regard. We reject similar criticisms of other precedent for the same reason.

3. Other Provisions of the Act Support Broadband's Information Service Classification

39. We also find that other provisions of the Act support our conclusion that broadband internet access service is best classified as an information service. We do not assert that the language in Sections 230 and 231 is determinative of the information service classification; rather, we find it to be supportive of our analysis of the textual provisions at issue. As such, we find Public Knowledge's assertions that the Commission's reasoning “would overrule the Supreme Court's holding in

Brand X

. . . [in which] the Court ruled that the Communications Act does not make explicit the correct classification of BIAS” inapposite. For instance, Congress codified its view in Section 230(b)(2) of the Act, stating that it is the policy of the United States “to preserve the vibrant and competitive free market that presently exists for the internet and other interactive computer services, unfettered by Federal or State regulation.” This statement confirms that the free market approach that flows from classification as an information service is consistent with Congress's intent. In contrast, we find it hard to reconcile this statement in Section 230(b)(2) with a conclusion that Congress intended the Commission to subject broadband internet access service to common carrier regulation under Title II.

40. Additional provisions within Sections 230 and 231 of the Act lend further support to our interpretation. Section 230(f)(2) defines an interactive computer service to mean “

any information service,

system, or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system

that provides access to the Internet

and such systems operated or services offered by libraries or educational institutions.” Thus, on its face, the plain language of this provision appears to reflect Congress' judgment that internet access service is an information service.

41. Section 230 states that an “information service” includes “a service or system that provides access to the internet,” and we disagree with commenters who read the definition of “interactive computer service” differently. Specifically, we disagree with commenters asserting that it is unclear whether the clause “including specifically a service . . . that provides access to the internet” modifies “information service” or some other noun phrase, such as “access software provider” or “system.” We think it a more reasonable interpretation that the phrase “service . . . that provides access to the internet” modifies the noun phrase “information service.” Similarly, we disagree that Section 230(f)(2) proves only “that there exist information services that provide access to the internet, not that all services that provide access to the internet are information services.” On the contrary, we agree with AT&T that “the formula `any X, including specifically a Y,' does logically imply that all Ys are Xs.”

42. Reliance on Section 230(f)(2) to inform the Commission's interpretations and applications of Titles I and II accords with widely accepted canons of statutory interpretation. The Supreme Court has recognized there is a “natural presumption that identical words used in different parts of the same act are intended to have the same meaning.” And there is nothing in the context of either section that overcomes the presumption. Indeed, the similarity of circumstances confirms the presumption of similar meaning, as the deregulatory approach to information services embodied in Titles I and II, as well as the deregulatory policy of Section 230, were all adopted as part of the 1996 Act. Thus, we disagree with the

Title II Order'

s argument that giving

Section 230 its plain meaning would be “an oblique” way to “settle the regulatory status of broadband internet access.” On the contrary, we agree that “it is hardly `oblique' for Congress to confirm in Section 230 that internet access should be classified as an unregulated information service when elsewhere in the same legislation Congress codifies a definition of `information services' that was long understood to include gateway services such as internet access.” And while the

USTelecom

court did not find this definition determinative on the issue, we find that “it is nonetheless a strong indicator that Congress was more comfortable with the prevailing view that provision of internet access is not a telecommunications service, and should not be subject to the array of Title II statutory provisions.” We find inapplicable the

USTelecom

court's invocation of the principle that “Congress . . . does not alter the fundamental details of a regulatory scheme in vague terms or ancillary provisions.” Section 230 did not alter any fundamental details of Congress's regulatory scheme but was part and parcel of that scheme, and confirmed what follows from a plain reading of Title I—namely, that broadband internet access service meets the definition of an information service. The legislative history of Section 230 also lends support to the view that Congress did not intend the Commission to subject broadband internet access service to Title II regulation. The congressional record reflects that the drafters of Section 230 did “not wish to have a Federal Computer Commission with an army of bureaucrats regulating the internet.” We likewise reject arguments premised on the theory that we are treating definitions in Section 230 and 231 as dispositive, rather than relying on them to inform our understanding of Congress' intent as revealed by the text and structure of the Act more broadly.

43. Section 231, inserted into the Communications Act a year after the 1996 Act's passage, similarly lends support to our conclusion that broadband internet access service is an information service. It expressly states that “internet access service” “does not include telecommunications services,” but rather “means a service that enables users to access content, information, electronic mail, or other services offered over the internet, and may also include access to proprietary content, information, and other services as part of a package of services offered to consumers.” Further, the carve-outs in Section 231(b)(1)-(2) differentiate the provision of telecommunications services and the provision of internet access service. It is hard to imagine clearer statutory language. The Commission has consistently held that categories of telecommunications service and information service are mutually exclusive; thus, because it is an information service, internet access cannot be a telecommunications service. Our interpretation of “telecommunications service” and “information service” as mutually exclusive ways to classify a given service thus demonstrates the relevance of Section 231 notwithstanding that it does not expressly define broadband internet access service as an information service. On its face then, this language strongly supports our conclusion that, under the best reading of the statute, broadband internet access service is an information service, not a telecommunications service. Nothing in the text of Section 231 reveals that the use of “internet access service” there is limited to dial-up internet access. To the contrary, it would seem anomalous for Congress only to exempt entities providing dial-up internet access and not other forms of internet access from the prohibitions of Section 231(a). We thus are unpersuaded by arguments advocating a narrower interpretation of “internet access service” in Section 231.

44. We also find that the purposes of the 1996 Act are better served by classifying broadband internet access service as an information service. Congress passed the Telecommunications Act to “promote competition and reduce regulation.” Further, as a bipartisan group of Senators stated, “[n]othing in the 1996 Act or its legislative history suggests that Congress intended to alter the current classification of internet and other information services or to expand traditional telephone regulation to new and advanced services.” Or as Senator John McCain put it, “[i]t certainly was not Congress's intent in enacting the supposedly pro-competitive, deregulatory 1996 Act to extend the burdens of current Title II regulation to internet services, which historically have been excluded from regulation.” It stands these goals on their head for the Commission, as deployment of advanced services reaches the mainstream of Americans' lives, to perpetuate the very Title II regulatory edifice that the 1996 Act sought to dismantle. An information service classification will “reduce regulation” and preserve a free market “unfettered by Federal or State regulation.”

45. Finally, we observe that the structure of Title II appears to be a poor fit for broadband internet access service. Indeed, numerous Title II provisions explicitly assume that all telecommunications services are a telephone service. For example, Section 221 addresses special provisions related to telephone companies, Section 251 addresses the obligations of local exchange carriers and incumbent local exchange carriers, and Section 271 addresses limitations on Bell Operating Companies' provision of interLATA services. For example, to obtain authority to offer in-region interLATA services, the BOCs have to offer a number of functions of particular relevance to the provision of telephone service. Therefore, it is no surprise that the

Title II Order

found that many provisions of Title II were ill-suited to broadband internet access services, and the Commission was forced to, on its own motion, forbear either in whole or in part on a permanent or temporary basis from 30 separate sections of Title II as well as from other provisions of the Act and Commission rules. We find that the significant forbearance the Commission deemed necessary in the

Title II Order

strongly suggests that the regulatory framework of Title II, which was specifically designed to regulate telephone services, is unsuited for the dissimilar and dynamic broadband internet access service marketplace.

B. Reinstating the Private Mobile Service Classification of Mobile Broadband Internet Access Service

46. Having determined that broadband internet access service, regardless of whether offered using fixed or mobile technologies, is an information service under the Act, we now address the appropriate classification of mobile broadband internet access service under Section 332 of the Act. We restore the prior longstanding definitions and interpretation of this section and conclude that mobile broadband internet access service should not be classified as a commercial mobile service or its functional equivalent.

47.

Background.

Section 332 of Title III, enacted by Congress as part of the Omnibus Budget Reconciliation Act of 1993 (the Budget Act), provides a specific framework that applies to providers of “commercial mobile service.” The section defines “commercial mobile service” as: “any mobile service . . . that is provided for profit and makes interconnected service available (A) to the public or (B) to such classes of eligible users as to be effectively available to a substantial portion of the public, as specified by regulation by the Commission.”

“Interconnected service,” in turn, is defined as “service that is interconnected with the public switched network (as such terms are defined by regulation by the Commission).” In 1994, the Commission adopted regulations implementing this section, codifying the definition of “commercial mobile service” under the term “commercial mobile radio service” (CMRS). Looking at the statute's text, structure, legislative history, and purpose, the Commission defined the “public switched network” as “[a]ny common carrier switched network, whether by wire or radio, including local exchange carriers, interexchange carriers, and mobile service providers, that use[s] the North American Numbering Plan in connection with the provision of switched services.” It defined “interconnected service” as “a service that gives subscribers the capability to communicate . . . [with]

all

other users on the public switched network.”

48. Section 332 distinguishes commercial mobile service from “private mobile service,” defined as “any mobile service . . . that is not a commercial mobile service or the functional equivalent of a commercial mobile service, as specified by regulation by the Commission.” In 1994, the Commission established its functional equivalence test, which starts with a presumption that “a mobile service that does not meet the definition of CMRS is a private mobile radio service.” Overcoming this presumption requires an analysis of a variety of factors to determine whether the mobile service in question is the functional equivalent of commercial mobile service, including “consumer demand for the service to determine whether the service is closely substitutable for a commercial mobile radio service; whether changes in price for the service under examination, or for the comparable commercial mobile radio service would prompt customers to change from one service to the other; and market research information identifying the targeted market for the service under review.” Emphasizing the high bar it had set, the Commission expected that “very few mobile services that do not meet the definition of CMRS will be a close substitute for a commercial mobile radio service.” We note that, in another Order adopted today, we are recodifying these factors under Section 20.3 of the Commission's rules, but not modifying their substance.

49. The Act treats providers of commercial mobile service as common carriers, and the legislative history of the 1996 Act suggests that Congress intended the definition of “telecommunications service” to include commercial mobile service. In contrast, the Act prohibits the Commission from treating providers of private mobile service as common carriers.

50. In 2007, the Commission found that wireless broadband internet access service was not a commercial mobile service because it did not meet the definition of an “interconnected service” under the Act and the Commission's rules. It found that wireless broadband internet access was not “interconnected” with the “public switched network” because it did not use the North American Numbering Plan, which limited “subscribers' ability to communicate to or receive communication from all users in the public switched network.” The Commission concluded that Section 332 and the Commission's rules “did not contemplate wireless broadband internet access service as provided today” and that a commercial mobile service “must still be interconnected with the local exchange or interexchange switched network as it evolves.”

51. In the

Title II Order,

the Commission reversed course. First, the Commission changed definitions of two key terms within the definition of commercial mobile service. It broadened the definition of the term “public switched network” to include services that use “public IP addresses.” And it redefined the term “interconnected service” by deleting the word “all” from the requirement that the service give subscribers the capability to communicate with “

all

other users on the public switched network,” so that a service would be interconnected even if users of such a service could not communicate with all other users. By manipulating these definitions, the Commission engineered a conclusion that mobile broadband internet access was interconnected with the public switched network and was an interconnected service under Section 332.

52. Second, the

Title II Order

found that even if it had not changed the definitions, it could change the scope of the service to meet them. Specifically, the Commission found that “users have the `capability' . . . to communicate with NANP numbers using their broadband connection through the use of VoIP applications.” Accordingly it found that, by including services not offered by the mobile broadband internet access service provider as part of the service, mobile broadband internet access service would now meet the regulatory definition of “interconnected service” adopted in 1994.

53. Third, the

Title II Order

eschewed the functional equivalence test contained in the Commission's rules to find that mobile broadband internet access service was functionally equivalent to commercial mobile service. Rather than apply that test, the Commission reasoned that the two were functionally equivalent because “like commercial mobile service, [mobile broadband internet access service] is a widely available, for profit mobile service that offers mobile subscribers the capability to send and receive communications on their mobile device to and from the public.”

54. In the

Internet Freedom Notice of Proposed Rulemaking (NPRM)

(82 FR 25568), the Commission proposed to “restore the meaning of `public switched network' under Section 332(d)(2) to its pre-

Title II Order

focus on the traditional public switched telephone network” and “to return to our prior definition of `interconnected service.' ” The Commission further proposed to return to the analysis of the

Wireless Broadband Internet Access Order

and find that mobile broadband internet access service was a private mobile service. Finally, it proposed to reconsider the

Title II Order'

s departure from the functional equivalence test codified in our rules.

55.

Discussion.

We find that the definitions of the terms “public switched network” and “interconnected service” that the Commission adopted in the 1994

Second CMRS Report and Order

reflect the best reading of the Act, and accordingly, we readopt the earlier definitions. We further find that, under these definitions, mobile broadband internet access service is not a commercial mobile service.

56. We find that the Commission's original interpretation of “public switched network” was more consistent with the ordinary meaning and commonly understood definition of the term and with Commission precedent. On multiple prior occasions before Section 332(d)(2) was enacted, the Commission used the term “public switched network” to refer to the traditional public switched telephone network. In 1981, for example, the Commission noted that “the public switched network interconnects all telephones in the country.” In 1992, the Commission described its cellular service policy as “encourag[ing] the creation of a nationwide, seamless system, interconnected with the public switched network so that cellular and landline telephone customers can

communicate with each other on a universal basis.” Courts also used the term “public switched network” when referring to the traditional telephone network. Based on this history of usage of the term, the Commission, in 1994, tied its definition of the term “public switched network” to the traditional switched telephone network. We find this approach appropriately reflects the fundamental canon of statutory construction that “unless otherwise defined, words will be interpreted as taking their ordinary, contemporary, common meaning.” We find that the legislative history of the Budget Act further supports this view. One commenter notes that the Budget Act conferees chose the Senate version of the relevant statutory definitions, including the use of the term “public switched network,” over the House version, which used the term “public switched telephone network,” and argues that Congress thereby rejected the latter term. We note, however, that the conferees also expressly identified the substantive differences between the House and Senate versions of the definitions, and notably absent from their list was any contrast between the Senate's use of “public switched network” and the House's use of “public switched telephone network,” suggesting that the conferees did not view the two terms as a significant difference.

57. We also find that the Commission's prior interpretation is more consistent with the text of Section 332(d)(2), in which Congress provided that commercial mobile service must provide a service that is interconnected with “

the

public switched network.” We find that the use of the definite article “the” and singular term “network” shows that Congress intended “public switched network” to mean a single, integrated network. We therefore agree with commenters who argue that it was not meant to encompass multiple networks whose users cannot necessarily communicate or receive communications across networks. Consistent with Congress's directive to define “

the

public switched network,” the restored definition reflects that the public switched network is a singular network that “must still be interconnected with the local exchange or interexchange switched network as it evolves,” as opposed to multiple networks that need not be connected to the public telephone network. That the Commission's original interpretation better reflects Congressional intent is further evidenced by the fact that, although Congress has amended the Communications Act and Section 332 on multiple occasions since the Commission defined the term, it has never changed the Commission's interpretation. As we further discuss elsewhere in connection with the term “interconnected service,” we find the best interpretation is to classify a service under Section 332 based solely on the nature of the service offered. Even if we were to consider such applications, however, we find that the public switched telephone network and the internet are and will continue to be distinct and separate networks, and cannot be considered a singular, integrated network as intended by the term “the public switched network.” The deployment of the Internet of Things (IoT), for example, will mean a dramatic increase in the number of non-VoIP-capable end-points, such as IP-enabled televisions, washing machines, and thermostats, and other smart devices.

58. We also restore the definition of “interconnected service” that existed prior to the

Title II Order.

Prior to that Order, the term was defined under the Commission's rules as a service “that gives subscribers the capability to communicate to or receive communication from

all

other users on the public switched network.” The

Title II Order

modified this definition by deleting the word “all,” finding that mobile broadband internet access service should still be considered an interconnected service even if it only enabled users to communicate with “some” other users of the public switched network rather than all. We agree with commenters who argue that the best reading of “interconnected service” is one that enables communication between its users and all other users of the public switched network. This reading ensures that the public switched network remains the single, integrated network that we find Congress intended in Section 332(d)(2), as reflected in the statutory definition of “interconnected service” as one that is interconnected with “the public switched network.” The

Title II Order

rejected this reading on the ground that the Commission has previously recognized that interconnected services may be limited in certain ways. While an interconnected service is required to provide its users with the capability to communicate with or receive communication from all other users of the public switched network, the Commission has permitted an interconnected service to restrict access to the public switched network in certain limited ways (such as the blocking of 900 numbers). This limited exception to general access has existed since the original definition of the term “interconnected service” was adopted, and the record does not demonstrate that it has caused confusion or misunderstandings about what services may be considered interconnected. Accordingly, we will continue to apply the definition of “interconnected service” in this fashion, and we see no need to codify any language further clarifying the exception. We agree with Verizon, however, that “[t]here is a massive difference between limited, targeted restrictions that deny access to certain points on the network and the situation envisioned by the

Title II Order,

where millions of users on what is ostensibly the same network are incapable of reaching each other.”

59. Some commenters who argue that the

Title II Order'

s revised definitions should be maintained point to Congress's delegation of interpretational authority to the Commission and the Commission's previous position that it could define the public switched network based on new technology and consumer demand. In defining the terms “public switched network” and “interconnected service” in the

Second CMRS Report and Order,

however, the Commission recognized that commercial mobile service must still be interconnected with the local exchange or interexchange switched network, and it stated that “any switched common carrier service that is interconnected with the traditional local exchange or interexchange switched network will be defined as part of that network for purposes of our definition of `commercial mobile radio services.' ” We disagree with commenters arguing that, by not including IP addresses in the definition of the public switched network, the Commission would be failing to recognize the evolution of mobile network technologies that have blurred the lines between circuit switched and packet switched networks. The Commission's original decision properly reflects that the public switched network should not be defined in a static way and should reflect that the public switched network is continuously growing and changing, but also ensures that, as it grows and evolves, the public switched network remains a single integrated network incorporating the traditional local and interexchange telephone networks and enabling users to send or receive messages to or from all other users. Further, although the

Title II Order

found that the revised definitions

adopted at that time were warranted as better reflecting current technological developments, including the “rapidly growing and virtually universal use of mobile broadband service” and the “universal access provided . . . by and to mobile broadband,” the Commission expressly noted that its determination was “a policy judgment that section 332(d) expressly delegated to the Commission, consistent with its broad spectrum management authority under Title III.” We find that this analysis places undue weight on the wide availability of a mobile service, as being effectively available to a substantial portion of the public is merely one of the definitional criteria. The Commission found that the updated definitions would be consistent with Congress's intent to create a symmetrical regulatory framework among mobile services that were similarly “broadly available” to the public. While we agree that Congress intended, in adopting Section 332, to regulate similar mobile services symmetrically, we do not believe that Congress intended for the Commission to regulate mobile services symmetrically simply because they are similarly “broadly available.” First, being “effectively available to a substantial portion of the public” is a necessary, but not sufficient, requirement for classification as commercial mobile service. Second, as noted, Congress set as the touchstone for regulatory symmetry only those mobile services that are “functionally equivalent.” In light of definitional analysis discussed above, as well as the public policy considerations that we have found to support our decision to classify broadband internet access service as an information service, we find under the same authority that such developments do not persuade us to retain the modified definitions.

60. We find that mobile broadband internet access service does not meet the regulatory definition of “interconnected service” that the Commission originally adopted in 1994 and which we readopt today, and therefore it does not meet the definition of commercial mobile service. As the Commission found in the

Wireless Broadband Internet Access Order,

“[m]obile wireless broadband Internet access service in and of itself does not provide the capability to communicate with all users of the public switched network” because it does “not use the North American Numbering Plan to access the Internet, which limits subscribers' ability to communicate to or receive communications from

all

users in the public switched network.” Accordingly, it is “not an `interconnected service' as the Commission has defined the term in the context of section 332.”

61. We disagree with the conclusion in the

Title II Order

that, because an end user can use a separate application or service that rides on top of the broadband internet access service for interconnected communications, mobile broadband internet access service meets the definition of “interconnected service.” We find that the definition of “interconnected service” focuses on the characteristics of the offered mobile service itself. Thus, the service in question must itself provide interconnection to the public switched network using the NANP to be considered an interconnected service. Our interpretation is consistent with Commission precedent that, prior to the

Title II Order,

had classified a service based on the nature of the service itself. This interpretation is also consistent with Section 332(d)(1), which defines commercial mobile service as a service that itself “makes interconnected service available . . . to the public,” and with Section 332(d)(2), which defines “interconnected service” as “service that is interconnected with the public switched network.” These statutory definitions focus on the functions of the service itself rather than “whether the service allows consumers to acquire other services that bridge the gap to the telephone network.” Thus, we are not persuaded by arguments that “applications such as Google Voice reflect the fully interconnected nature of the mobile broadband and legacy telephone networks.” Our determination reflects that the relevant service must itself be an “interconnected service,” and not merely a capability to acquire interconnection. We further note that viewing broadband internet access service as a distinct service from application layer services that may be accessed by it, even if the applications are pre-installed in the mobile device offered by the provider, ensures that similar mobile broadband internet access services are not regulated in a disparate fashion based on what applications a particular provider chooses to install in their offered devices. This is consistent with the fundamental purpose under Section 332 of regulatory symmetry between similar mobile services, and also avoids regulatory inconsistencies that would result when mobile devices are brought to a particular service provider by the consumer that do not include the provider's choice of pre-installed apps. While OTI New America argues that the need to obtain such apps to make an interconnected call does not make mobile broadband internet access service different from traditional telephone service, which has always required customer premises equipment to complete an interconnected call, we find the analogy inapt. With traditional CMRS, even where consumers obtain their premises equipment or mobile devices separately, the function of interconnection is provided by the purchased mobile service itself. Because the focus is solely on the relevant service provided, we also disagree that physical connections between networks, in and of themselves, establish that the relevant services are interconnected, and we further disagree that mobile broadband internet access service should be considered an interconnected service simply because a separate interconnected voice service may be provided using the same packet-switched network layer.

62. Consistent with the Commission's analysis in the

Wireless Broadband Internet Access Order,

the fact that “consumers are now able to use a variety of Internet-enabled applications that allow them to send calls and texts to NANP end-points” does not make mobile broadband internet access service itself an interconnected service as defined by our rules. The increased use and availability of mobile VoIP applications does not change the fact that mobile broadband internet access as a core service is distinct from the service capabilities offered by applications (whether installed by a user or hardware manufacturer) that may ride on top of it. When viewed as a distinct service, it is apparent that today's mobile broadband internet access service itself does not enable users to reach NANP telephone numbers and therefore cannot be considered an interconnected service. We do not here address whether IP-based services or applications such as Wi-Fi Calling or VoLTE would meet the definition of “interconnected service” under Section 332 and the Commission's rules. We disagree with OTI New America's argument that the growing availability of Wi-Fi Calling provided by mobile carriers that also offer mobile broadband internet access service supports the classification of mobile broadband internet access service as a commercial mobile service. The two are distinct services and subject to separate classification determinations. Similarly, even if providers are increasingly offering voice service and mobile broadband internet access service

together, this does not support classifying and regulating the latter in the same way as the former. Providers have long offered multiple services of mixed classification, subject to the rule that they are regulated as common carriers to the extent they offer services that are subject to Title II regulation.

63. Moreover, in light of the determination above that mobile broadband internet access service should be restored to its classification as an information service, and consistent with our findings today that reinstating this classification will serve the public interest, we also find that it will serve the public interest for the Commission to exercise its statutory authority to return to its original conclusion that mobile broadband internet access is not a commercial mobile service. We note that commenters who support the

Title II Order's

revised definition of “public switched network” do not dispute that Congress expressly delegated authority to the Commission to define the key terms,

i.e.,

“public switched network” and “interconnected service.” No one disputes that, consistent with the Commission's previous findings, if mobile broadband internet access service were a commercial mobile service for purposes of Section 332 and were also classified as an information service, such a regulatory framework could lead to contradictory and absurd results. Among these problems, as the Commission explained in 2007, is that a contrary reading of the Act would result in an internal contradiction within the statutory framework, because Section 332 would require that the service provider be treated as a common carrier insofar as it provides mobile wireless broadband internet access service, while Section 3 clearly would prohibit the application of common carrier regulation of such a service provider's provision of that service. Indeed, the

Title II Order,

like the 2007

Wireless Broadband Internet Access Order,

recognized and sought to avoid the significant problems in construing Section 332 in a manner that set up this “statutory contradiction” with the scope of Title II. Construing the CMRS definition to exclude mobile broadband internet access service as an information service similarly avoids this contradiction, furthers the Act's overall intent to allow information services to develop free from common carrier regulations, and is consistent with the public policy analysis in connection with our determination to reclassify mobile broadband internet access as an information service. Further, it avoids the absurd result of singling out mobile providers of broadband internet access service for such common carrier regulation while freeing fixed broadband internet access services from such regulation, notwithstanding that, as discussed elsewhere in this Order, there is generally greater competition in the provision of mobile broadband internet access service than in fixed broadband internet access service. We note that wireless services similar to mobile broadband internet access service were not available in the market place in 1993 when Congress adopted Section 332 or, in 1996, when Congress adopted the Section 3 definition of “telecommunication carrier.”

64. In addition to finding that mobile broadband internet access is not a commercial mobile service, we also adopt our proposal to reconsider the Commission's analysis regarding functional equivalence in the

Title II Order.

For the same reasons discussed below with respect to our authority to revisit the classification of broadband internet access service, we disagree with arguments regarding limits on the Commission's ability to revisit the

Title II Order'

s findings regarding functional equivalence. In addition, we note that the

Title II Order,

in reaching the conclusion that mobile broadband internet access was a commercial mobile service, relied in part on the need to avoid a statutory contradiction with its determination that the service was a telecommunications service. Given our decision to restore the original classification of mobile broadband internet access service as an information service, this change additionally warrants revisiting our conclusions with regard to the classification of mobile broadband internet access service under Section 332. We find that the test for functional equivalence adopted in the

Second CMRS Report and Order

reflects the best interpretation of Section 332. Under this test, a variety of factors will be evaluated to make a determination whether the mobile service in question is the functional equivalent of a commercial mobile radio service, including: Consumer demand for the service to determine whether the service is closely substitutable for a commercial mobile radio service; whether changes in price for the service under examination, or for the comparable commercial mobile radio service would prompt customers to change from one service to the other; and market research information identifying the targeted market for the service under review. In contrast, as noted above, the

Title II Order

based its finding of functional equivalence on the notion that “like commercial mobile service, [mobile broadband Internet access] is a widely available, for profit mobile service that offers mobile subscribers the capability to send and receive communications on their mobile device to and from the public.” Commenters who support the classification of mobile broadband internet access service as a commercial mobile service similarly contend that mobile broadband internet access service shares no similarities with other private mobile services such as taxi dispatch services and that, in contrast, “there is no networked service more open, interconnected, and universally offered than mobile broadband Internet access service.” We note that the statute directs us to determine whether mobile broadband internet access is functionally equivalent to a commercial mobile service, not whether it is functionally dissimilar from certain systems classified as private mobile.

65. We believe the test of functional equivalence adopted in the

Second CMRS Report and Order

hews much more faithfully to the intent of Congress than the approach applied in the

Title II Order

or the analyses in the record focusing on the extent of service availability. If Congress meant for widespread public access to a widely used service to be the determining factor for what is “functionally equivalent” to a commercial mobile service, it would not have included being “interconnected with the public switched network” in the statutory definition of the service. Indeed, the relevant House Report, in describing “private carriers” that under the current law were offering service “[f]unctionally . . . indistinguishable” from carriers classified as common carriers, highlighted that these private carriers were offering services interconnected with the public switched network. Although the Commission has discretion to determine whether services are functionally equivalent, we find that the

Title II Order'

s reliance on the public's “ubiquitous access” to mobile broadband internet access service alone was insufficient to establish functional equivalency. In contrast, the test established in the

Second CMRS Report and Order

provides a thorough consideration of factors that are indicative of whether a service is closely substitutable in the eyes of consumers for a commercial mobile service.

66. Applying the test adopted by the Commission in the

Second CMRS Report and Order,

we find that mobile broadband internet access service today is not the functional equivalent of

commercial mobile service as defined by the Commission. We note again that, under this test, services not meeting the definition of commercial mobile service are presumed to be not functionally equivalent, a presumption particularly intuitive here in light of the functional differences between traditional commercial mobile services like mobile voice and today's mobile broadband services. The evidence on demand substitutability only reinforces this presumption. First, mobile broadband internet access service and traditional mobile voice services have different service characteristics and intended uses. Consumers purchase mobile broadband internet access service to access the internet, on-line video, games, search engines, websites, and various other applications, while they purchase mobile voice service solely to make calls to other users using NANP numbers. Pricing and marketing information similarly support the conclusion that today mobile broadband internet access service and traditional mobile voice services are not “closely substitutable.” Such evidence suggests, for example, that mobile service providers target different types of customer groups when advertising voice, as opposed to mobile broadband internet access service. Moreover, at this time, voice-only mobile services tend to be much less expensive than mobile broadband internet access services, and they appear to be targeted to consumers who seek low-cost mobile service. Currently, for example, unlimited voice and text only plans may range from $15 to $25 per month. In contrast, unlimited mobile broadband internet plans may range from $60 to $90 per month for a single line. Nothing in the record suggests that changing the price for one service by a small but significant percentage would prompt a significant percentage of customers to move to the other service. Accordingly, under the functional equivalence standard adopted in the

CMRS Second Report and Order,

we find that mobile broadband internet access today is not the functional equivalent of commercial mobile service. The two services have different service characteristics and intended uses and are not closely substitutable for each other, as evidenced by the fact that changes in price for one service generally will not prompt significant percentages of customers to change from one service to the other. We make a conforming revision to the definition of “commercial mobile radio service” in Section 20.3 of the Commission's rules to reflect our determination that mobile broadband internet access service is not the functional equivalent of commercial mobile service.

C. Public Policy Supports Classifying Broadband Internet Access Service as an Information Service

67. While our legal analysis concluding that broadband internet access service is best classified as an information service under the Act is sufficient grounds alone on which to base our classification decision, the public policy arguments advanced in the record and economic analysis reinforce that conclusion. We find that reinstating the information service classification for broadband internet access service is more likely to encourage broadband investment and innovation, furthering our goal of making broadband available to all Americans and benefitting the entire internet ecosystem. For almost 20 years, there was a bipartisan consensus that broadband should remain under Title I, and ISPs cumulatively invested $1.5 trillion in broadband networks between 1996 and 2015. Commenters who claim recent growth in online video streaming services is evidence of the need for Title II regulation ignore the fact that the growth of online video streaming services was largely made possible by the network investments made under Title I and as such demonstrates instead the success of the longstanding light-touch framework under Title I. During that period of intense investment, broadband deployment and adoption increased dramatically, as the combined number of fixed and mobile internet connections increased from 50.2 million to 355.2 million from 2005 to 2015, and even as early as 2011, a substantial majority of Americans had access to broadband at home. As of 2016, roughly 91 percent of homes had access to networks offering 25 Mbps, and there were 395.9 million wireless connections, twenty percent more than the U.S. population. Mobile data speeds have also dramatically increased, with speeds increasing 40-fold from the 3G speeds of 2007. Cable broadband speeds increased 3,200 percent between 2005 and 2015, while prices per Mbps fell by more than 87 percent between 1996 and 2012.

68. Based on the record in this proceeding, we conclude that economic theory, empirical studies, and observational evidence support reclassification of broadband internet access service as an information service rather than the application of public-utility style regulation on ISPs. We find the Title II classification likely has resulted, and will result, in considerable social cost, in terms of foregone investment and innovation. At the same time, classification of broadband internet access service under Title II has had no discernable incremental benefit relative to Title I classification. The regulations promulgated under the Title II regime appear to have been a solution in search of a problem. Close examination of the examples of harm cited by proponents of Title II to justify heavy-handed regulation reveal that they are sparse and often exaggerated. Moreover, economic incentives, including competitive pressures, support internet openness. We find that the gatekeeper theory, the bedrock of the

Title II Order'

s overall argument justifying its approach, is a poor fit for the broadband internet access service market. Further, even if there may be potential harms, we find that pre-existing legal remedies, particularly antitrust and consumer protection laws, sufficiently address such harms so that they are outweighed by the well-recognized disadvantages of public utility regulation. As such, we find that public policy considerations support our legal finding that broadband internet access service is an information service under the Act.

1. Title II Regulation Imposes Substantial Costs on the Internet Ecosystem

69. The Commission has long recognized that regulatory burdens and uncertainty, such as those inherent in Title II, can deter investment by regulated entities and, until the

Title II Order,

its regulatory framework for cable, wireline, and wireless broadband internet access services reflected that reality. Congress has similarly recognized the burdens associated with regulation. For example, the 1996 Act states its purpose is to “reduce regulation,” and directs the Commission to regularly review regulations and repeal those it deems unnecessary or harmful to investment, competition, and the public interest. This concern is well-documented in the economics literature on regulatory theory, and the record also supports the theory that the regulation imposed by Title II will negatively impact investment. The balance of the evidence in the record suggests that Title II classification has reduced ISP investment in broadband networks, as well as hampered innovation, because of regulatory uncertainty. The record also demonstrates that small ISPs, many of which serve rural consumers, have been particularly harmed by Title II. And there is no convincing evidence of increased investment in the edge that

would compensate for the reduction in network investment.

70.

Investment by ISPs.

As the Commission has noted in the past, increased broadband deployment and subscribership require investment, and the regulatory climate affects investment. The mechanisms by which public utility regulation can depress investment by the regulated entity are well-known in the regulatory economics literature. The owners of network infrastructure make long-term, irreversible investments. In theory, public utility regulation is intended to curb monopoly pricing just enough that the firm earns a rate of return on its investments equivalent to what it would earn in a competitive market. In practice, public utility regulation can depress profits below the competitive rate of return for a variety of reasons. This reduction in the expected return reduces the incentive to invest. Importantly, the risk that regulation might push returns below the competitive level also creates a disincentive for investment.

71. We first look to broadband investment in the aggregate and find that it has decreased since the adoption of the

Title II Order.

ISP capital investment increased each year from the end of the recession in 2009 until 2014, when it peaked. In 2015, capital investment by broadband providers appears to have declined for the first time since the end of the recession in 2009. And investment levels fell again in 2016—down more than 3 percent from 2014 levels. Although declines in broadband capital investments have occurred in the past with changes in the business cycle, the most recent decline is particularly curious given that the economy has not experienced a recession in recent years but rather has been growing. While observing trends in the data by itself cannot establish the cause of directional movements, the stark trend reversal that has developed in recent years suggests that changes to the regulatory environment created by the

Title II Order

have stifled investment. In addition to data trends, the record contains a variety of other studies, using different methodologies which seek to determine how imposition of public-utility style regulation might affect ISPs' investments.

72. Comparisons of ISP investment before and after the

Title II Order

suggest that reclassification has discouraged investment. Performing such a comparison, economist Hal Singer concluded that ISP investment by major ISPs fell by 5.6 percent between 2014 and 2016. Singer attempted to account for a few significant factors unrelated to Title II that might affect investment, by subtracting some investments that are clearly not affected by the regulatory change (such as the accounting treatment of Sprint's telephone handsets, AT&T's investments in Mexico, and DirecTV investments following its acquisition by AT&T in the middle of this period). In contrast, Free Press presents statistics that it claims demonstrate that broadband deployment and ISP investment “accelerated” to “historic levels” after the Commission approved the

Title II Order.

But Free Press fails to account for factors such as foreign investment and the appropriate treatment of handsets as capital expenditures, as Singer did.

73. A comparative assessment that adjusted the Free Press and Singer numbers so that they covered the same ISPs, spanned the same time period, and subtracted investments unaffected by the regulatory change, found that both sets of numbers demonstrate that ISP investment fell by about 3 percent in 2015 and by 2 percent in 2016. A Free State Foundation calculation using broadband capital expenditure data for 16 of the largest ISPs reached a result similar to Singer's, but this analysis simply compared actual ISP investment to a trend extrapolated from pre-2015 data. These types of comparisons can only be regarded as suggestive, since they fail to control for other factors that may affect investment (such as technological change, the overall state of the economy, and the fact that large capital investments often occur in discrete chunks rather than being spaced evenly over time), and companies may take several years to adjust their investment plans. Nonetheless, these comparisons are consistent with other evidence in the record that indicates that Title II adversely affected broadband investment. A separate comparison of the United States' ISP investment with ISP investment in Europe also suggests that ISP investment might decline further if the U.S., under the

Title II Order,

moves toward a regulatory system more like Europe's. A USTelecom research brief finds that European investment per capita is about 50 percent lower than broadband investment in the U.S. per capita. As some commenters point out, this study compares the U.S. with the much more regulatory European system, which includes mandatory unbundling at regulated rates. Thus, it presents a picture of how investment could change if the U.S. moves toward the European system under Title II, not an assessment of the direct results of the

Title II Order.

74. The record also contains analyses attempting to assess the predicted causal effects of Title II regulation on ISP investment and/or output. Some of these studies are “natural experiments” that seek to compare outcomes occurring after policy changes to a relevant counterfactual that shows what outcomes would have occurred in the absence of the policy change. No single study is dispositive, but methodologies designed to estimate impacts relative to a counterfactual tend to provide more convincing evidence of causal impacts of Title II classification. Having reviewed the record of these studies, the balance of the evidence indicates that Title II discourages investment by ISPs—a finding consistent with economic theory. The record does not provide sufficient evidence to quantify the size of the effect of Title II on investment. An additional type of evidence is the effect of the

Title II Order

on stock prices. According to that study, in the short term, the decision appears to have had little direct effect on stock prices, except for a few cable ISPs. That may reflect the forward-looking, predictive capabilities of market players.

75. Prior FCC regulatory decisions provide a natural experiment allowing this question to be studied. Scholars employing the natural experiment approach found that prior to 2003, subscribership to cable modem service (not regulated under Title II) grew at a far faster rate than subscribership to DSL internet access service (the underlying `last mile' facilities and transmission which were regulated under Title II). After 2003, when the Commission removed line-sharing rules on DSL, DSL internet access service subscribership experienced a statistically significant upward shift relative to cable modem service. A second statistically significant upward shift in DSL internet access service subscribership relative to cable modem service occurred after the Commission classified DSL internet access service as an information service in 2005. This evidence suggests that Title II discourages not just ISP investment, but also deployment and subscribership, which ultimately create benefits for consumers. While some commenters contend that deployment and subscribership continued to increase after the

Title II Order,

such that nothing is amiss, this casual observation does not compare observed levels of subscribership and deployment to a relevant counterfactual that controls for other factors.

76. An assessment of how ISP investment reacted to news of impending Title II regulation suggests that the threat of Title II regulation discouraged ISP investment. Such statistical analysis allows one to compare the actual level of investment with a counterfactual estimate of what investment would have been in the absence of the change in risk. This study found that Chairman Genachowski's 2010 announcement of a framework for reclassifying broadband under Title II—a credible increase in the risk of reclassification that surprised financial markets—was associated with a $30 billion-$40 billion annual decline in investment in the U.S. Bureau of Economic Analysis' “broadcasting and telecommunications” category between 2011 and 2015. The study attributes the decline to the threat of Title II regulation, rather than net neutrality

per se,

because no similar decline occurred when the FCC adopted the four principles to promote an open internet in 2005. Because the study's measure of investment data covers the entire broadcasting and telecommunications industries, the change in investment measured in this study might be larger than the change in broadband investment associated with the threat of Title II regulation. Accordingly, the findings may be a more reliable indicator of the direction of the change in investment than the absolute size of the change. At the very least, the study suggests that news of impending Title II regulation is associated with a reduction in ISP investment over a multi-year period.

77. Some commenters have argued that this study does not identify the effect of Title II on ISP investment, because the “last mile” facilities and transmission underlying DSL internet access service (essentially incumbent LEC broadband supply) were under Title II before 2005, during the study's pre-treatment period. However, to the extent that a fraction of the industry was subject to Title II (and at the time the bulk of broadband subscribers used cable modem services that were not regulated under Title II), this would imply Ford's negative result for investment was understated.

78. The study is also disputed by the Internet Association, which submitted an economic study arguing that the threat and eventual imposition of Title II status on broadband internet service providers in 2010 and 2015 did not have a measurable impact on telecommunications investment in the U.S. While we appreciate the alternative method and data sources introduced by that study, several elements lead us to discount its findings. The estimation of the impact of events in both 2010 and 2015 relies partially on forecast rather than actual data, which likely lessens the possibility of finding an effect of Title II on investment. In addition, when examining cable and telecommunications infrastructure investment in the U.S., the study relies on a regression discontinuity over time model, thereby eliminating the use of a separate control group to identify the effect of policy changes. We believe use of such a model in these circumstances is unlikely to yield reliable results. The Internet Association study claims that its test of the 2010 effect did not use forecast data. However, comparing the reported number of observations in Tables B1 and B2 of the study clearly indicates that the same datasets were used to estimate 2010 and 2015 effects. Furthermore, we note that the Phoenix Center attempted to replicate the results of Table B1 and obtained strikingly different results when excluding the forecast data. Unfortunately, the Phoenix Center chose to only estimate Hooton's baseline model, which did not control for obviously confounding factors such as the business cycle, and therefore we place limited weight on the Phoenix Center's revisions.

79. In light of the foregoing record evidence, we conclude that reclassification of broadband internet access service from Title II to Title I is likely to increase ISP investment and output. The studies in the record that control the most carefully for other factors that may affect investment (the Ford study and the Hazlett & Wright study) support this conclusion. Ford controls for macroeconomic factors that influence the overall economy using a two-way fixed-effects model. Hazlett & Wright's analysis of the effects of Title II on DSL subscribership cites regression analysis that controls for factors influencing the overall economy by including Canadian DSL subscribership as an explanatory variable. Consequently, we disagree with commenters who assert that Title II has increased or had no effect on ISP investment, given the failure of other studies to account for complexity of corporate decision-making and the macroeconomic effects that can play a role in investment cycles. We also disagree with commenters who assert that it may be too soon to meaningfully assess the economic effects that Title II has had on broadband infrastructure investment.

80.

Regulatory Uncertainty.

The evidence that Title II has depressed broadband investment is bolstered by other record evidence showing that Title II stifled network innovation. Among the unseen social costs of regulation are those broadband innovations and developments that never see the light of day. ISP investment does not simply take the form of greater deployment, but can also be directed toward new and more advanced services for consumers. Research and development is an inherently risky part of any business, and the Commission's actions should not introduce greater uncertainty and risk into the process without a clear need to do so. Numerous commenters have stated that the uncertainty regarding what is allowed and what is not allowed under the new Title II broadband regime has caused them to shelve projects that were in development, pursue fewer innovative business models and arrangements, or delay rolling out new features or services. Even large ISPs with significant resources have not been immune to the dampening effect that uncertainty can have on a firm's incentive to innovate. Charter, for instance, has asserted that it has “put on hold a project to build out its out-of-home Wi-Fi network, due in part to concerns about whether future interpretations of Title II would allow Charter to continue to offer its Wi-Fi network as a benefit to its existing subscribers.” Cox has also stated that it has approached the “development and launch of new products and service features with greater caution” due to the uncertainty created by the Title II classification. And while new service offerings can take a while to develop and launch, Comcast cites “Title II overhang” as a burden that delayed the launch of its IP-based transmission of its cable service, due to a year-long investigation.

81. Utility-style regulation is particularly inapt for a dynamic industry built on technological development and disruption. It is well known that extensive regulation distorts production as well as consumption choices. Regulated entities are inherently restricted in the activities in which they may engage, and the products that they may offer. Asking permission to engage in new activities or offer new products or services quickly becomes a major preoccupation of the utility. This is apparent upon a casual observation of heavily-regulated utilities, such as the U.S. power, water, and mass transit systems. These are industries where competition has been effectively deemed impossible, run by quasi-public monopolies that lack incentives to invest, innovate, or even properly maintain their facilities.

Within the communications industry, it is apparent that the most regulated sectors, such as basic telephone service, have experienced the least innovation, whereas those sectors that have been traditionally free to innovate, such as internet service, have greatly evolved. In the communications industry, incumbents have often used Commission regulation under the direction of the “public interest” to thwart innovation and competitive entry into the sector and protect existing market structures. Given the unknown needs of the networks of the future, it is our determination that the utility-style regulations potentially imposed by Title II run contrary to the public interest.

82. The record confirms that concern about “regulatory creep”—whereby a regulator slowly increases its reach and the scope of its regulations—has exacerbated the regulatory uncertainty created by the

Title II Order.

Even at the time of adoption, the Commission itself did not seem to know how the

Title II Order

would be interpreted. As then-Chairman Wheeler stated in February 2015, “we don't really know. No blocking, no throttling, no fast lanes. Those can be bright-line rules because we know about those issues. But we don't know where things go next.” With future regulations open to such uncertainties, Title II regulation adds a risk premium on each investment decision, which reduces the expected profitability of potential investments and deters investment. For example, the

Title II Order

did not forbear from

ex post

enforcement actions related to subscriber charges, raising concerns that

ex post

price regulation was very much a possibility. Further, providers have asserted that although the Commission forbore from the full weight of Title II in the

Title II Order,

they were less willing to invest due to concerns that the Commission could reverse course in the future and impose a variety of costly regulations on the broadband industry—such as rate regulation and unbundling/open access requirements—placing any present investments in broadband infrastructure at risk. These concerns were compounded by the fact that while the

Title II Order

itself announced forbearance from

ex ante

price regulation, at the same time it imposed price regulation with its ban on paid prioritization arrangements, which mandated that ISPs charge edge providers a zero price. These threats to the ISP business model have been felt throughout financial markets. As Craig Moffett of MoffettNathanson explained, “[i]t would be naïve to suggest that the implication of Title II, particularly when viewed in the context of the FCC's repeated findings that the broadband market is non-competitive, doesn't introduce a real risk of price regulation.” These risks are not merely theoretical: As CenturyLink contends, financial analysts lowered industry stock ratings due in part to the major risks Title II posed to the industry, which resulted in lower stock prices and lost market capitalization.

83. For these reasons, “any rational ISP will think twice before investing in innovative business plans that might someday be found to violate the Commission's undisclosed policy preferences and thus give rise to a cease-and-desist order and perhaps massive forfeiture penalties.” We conclude that this ever-present threat of regulatory creep is substantially likely to affect the risk calculus taken by ISPs when deciding how to invest their shareholders' capital, potentially deterring them from investing in broadband, and to encourage them to direct capital toward less inherently-risky business operations. Many ISPs are part of integrated multi-sector holding companies, which allows them to more easily shift capital away from sectors where their investments would face greater regulatory risk, and toward more investment-friendly sectors. We find unpersuasive the alleged inconsistencies between ISPs claiming that the

Title II Order

decreased their willingness or ability to invest in broadband infrastructure, and their statements to investors that the

Title II Order

has not had a negative impact on their broadband deployments. First, some of the comments claiming that corporate officers' statements to investors prove that Title II has increased investment use highly selective quotations that ignore other statements to investors that imply the opposite. Second, as other commenters point out, the latter often constitute statements susceptible to multiple interpretations, such as AT&T CEO Randall Stephenson stating that his company planned to “deploy more fiber next year than [it] did this year.” Third, these ambiguous statements do not take into account the relevant counterfactual scenario in which Title II regulation had not been adopted. Fourth, we observe that some of the comments attempting to highlight a discrepancy between statements to investors and statements in this proceeding simply show executives stating that their business

practices

will not change because they were not engaged in the conduct prohibited by the

Title II Order,

not that the firms' investment priorities remained the same after the

Title II Order.

As such, we disagree with commenters who assert that maintaining the

Title II Order

regime is the best means of addressing regulatory uncertainty.

84.

Small ISPs and Rural Communities.

The Commission's decision in 2015 to reclassify broadband internet access service as a telecommunications service has had particularly deleterious effects on small ISPs and the communities they serve, which are often rural and/or lower-income. The record reflects that small ISPs and new entrants into the market face disproportionate costs and burdens as a result of regulation. Many small ISPs lack the extensive resources necessary to comply with burdensome regulation, and the record evinces a widespread consensus that reclassification of broadband internet access service as a telecommunications service has harmed small ISPs by forcing them to divert significant resources to legal compliance and deterring them from taking financial risks.

85. Small ISPs state that these increased compliance costs and regulatory burdens have forced them to divert money and attention away from planned broadband service and network upgrades and expansions, thus delaying, deferring, or forgoing the benefits they would have brought “to their bottom lines, their customers, and their communities.” A coalition of National Multicultural Organizations highlights that the uncertainty inherent under Title II “already has produced results that slow needed innovation and broadband adoption, effects that are most acutely felt in rural and socioeconomically-challenged urban communities.” The record is replete with instances in which small ISPs reduced planned, or limited new, investment in broadband infrastructure as a result of the regulatory uncertainty stemming from the adoption of the

Title II Order.

Because the logical expectation that Title II regulation would have particularly harmful effects on small ISPs and the communities they serve in is borne out by strong record evidence from a wide range of small ISPs, we are unpersuaded by speculative suggestions that small ISPs' investment decisions can be fully or primarily explained based on other considerations such that the effect of Title II regulation can be neglected. The Wireless Internet Service Providers Association (WISPA) surveyed its members and found that over 80 percent had “incurred additional expense in complying with

the Title II rules, had delayed or reduced network expansion, had delayed or reduced services and had allocated budget to comply with the rules.” The threat of

ex post

rate regulation has hung particularly heavily on the heads of small ISPs, “who are especially risk-averse, causing them to run all current and planned offerings against the `just' and `reasonable' and unreasonably discriminatory standards of sections 201 and 202 of the Act.” The effects have been strongly felt by small ISPs, given their more limited resources, leading to depressed hiring in rural areas most in need of additional resources.

86. Compounding the difficulties faced by small ISPs, the record also reflects that the “ `black cloud' of common carriage regulations” resulted in increased difficulties for small ISPs in obtaining financing. A coalition of 70 small wireless ISPs cited the uncertainty created by the

Title II Order

as a major reason that their costs of capital have risen, preventing them from further expanding and improving their networks. The new regulatory burdens, risks, and uncertainties combined with “diminished access to capital create a vicious cycle—the regulatory burdens make it more difficult to attract capital, and less capital makes it more difficult to comply with regulatory burdens.” A coalition of 19 municipal ISPs cited high legal and consulting fees necessary to navigate the

Title II Order,

as well as regulatory compliance risk as a reason for delaying or abandoning new features and services. While, of course, not all small ISPs have faced these challenges, there is substantial record evidence that regulatory uncertainty resulting from the Commission's reclassification of broadband internet access service in 2015 risks stifling innovation, and that it has already done so with respect to small ISPs, which ultimately harms consumers.

87. We anticipate that the beneficial effects of our decision today to restore the classification of broadband internet access service to an information service will be particularly felt in rural and/or lower-income communities, giving smaller ISPs a stronger business case to expand into currently unserved areas. Enabling ISPs to freely experiment with services and business arrangements that can best serve their customers, without excessive regulatory and compliance burdens, is an important factor in connecting underserved and hard-to-reach populations. We are committed to bridging the digital divide, and recognize that small ISPs “disproportionately provide service in rural and underserved areas where they are either the only available broadband service option or provide the only viable alternative to an incumbent broadband provider.” We anticipate that returning broadband internet access service to a light-touch regulatory framework will help further the Commission's statutory imperative to “encourage the deployment on a reasonable and timely basis of advanced telecommunications capability to all Americans” by helping to incentivize ISPs to expand coverage to underserved areas. We therefore reject arguments that our classification decision harms low-income communities.

88.

Investment at the Edge.

Finally, to more fully discern the impact of Title II, we must look at investment throughout the broadband ecosystem, including investment and innovation at the edge, as well as with other ecosystem participants (manufacturers, etc.). We agree with commenters who assert that looking only at ISP investment ignores investment that is occurring at the edge. While there is tremendous investment occurring at the edge, the record does not suggest a correlation between edge provider investment and Title II regulation, nor does it suggest a causal relationship that edge providers have increased their investments as a result of the

Title II Order.

Free Press argues that since adoption of the

Title II Order,

innovation and investment at the edge has increased. While high growth rates are associated with the internet industry, the evidence presented does not show the imposition of Title II regulation on internet access service providers caused recent edge provider investment. That requires an estimate as to what would have happened in the absence of Title II regulation (

e.g.,

analysis following the methods employed in the studies of Ford, and of Hazlett & Wright).

89. In fact, one could argue that in the absence of Title II regulation, edge providers would have made even higher levels of investment than they undertook. In many cases, the strongest growth for a firm or industry predates the

Title II Order.

For example, Free Press highlights that the data processing, hosting, and related services industry increased capital expenditures by 26 percent in 2015, a significant increase in investment. However, in 2013, well before the 2014

Open Internet NPRM

that led to the

Title II Order,

that industry increased investment by over 100 percent. Similarly, Netflix's greatest relative increase in capital expenditures occurred in 2013. Amazon increased its spending on technology and content, which consists primarily of research and development expenses, by 28 percent in 2016, while in 2013 the increase was 41 percent. We do not claim that these data points prove that edge provider investment would have been greater in the absence of the

Title II Order,

but we find that Free Press does not demonstrate that there is a significant difference in the investment behavior of edge providers due to the

Title II Order.

2. Utility-Style Regulation of Broadband Is a Solution in Search of a Problem

90.

The internet was open before Title II, and many economic factors support openness.

The internet thrived for decades under the light-touch regulatory regime in place before the

Title II Order,

as ISPs built networks and edge services were born. We find that the sparse evidence of harms discussed in the

Title II Order

—evidence repeated by commenters in this proceeding as the basis for adopting a Title II classification—demonstrates that the incremental benefits of Title II over light-touch regulation are inconsequential, and pale in comparison to the significant costs of public-utility regulation. We therefore reject the argument that sparse evidence of harms is sufficient to justify the imposition of Title II.

91. The internet as we know it developed and flourished under light-touch regulation. It is self-evident that the hypothetical harms against which the

Title II Order

purported to protect did not thwart the development of the internet ecosystem. Edge providers have been able to disrupt a multitude of markets—finance, transportation, education, music, video distribution, social media, health and fitness, and many more—through innovation, all without subjecting the networks that carried them to onerous utility regulation. It is telling that the

Title II Order

and its proponents in this proceeding can point only to a handful of incidents that purportedly affected internet openness, while ignoring the two decades of flourishing innovation that preceded the

Title II Order.

92. The first instance of actual harm cited by the

Title II Order

involved Madison River Communications, a small DSL provider accused in 2005 of blocking ports used for VoIP applications, thereby foreclosing competition to its telephony business. Madison River entered into a consent decree with the Enforcement Bureau, paying $15,000 to the U.S. Treasury and agreeing that it “shall not block ports used for VoIP applications or otherwise prevent customers from using VoIP applications.” Vonage, an over-the-top

VoIP provider, later confirmed in press reports that it had initiated a complaint against Madison River at the Commission and that other small ISPs had blocked its VoIP services.

93. Next, the

Title II Order

referenced Comcast's throttling of BitTorrent, a peer-to-peer networking protocol. Comcast, which was at the time the nation's second-largest ISP, admitted that it interfered with about a tenth of BitTorrent TCP connections, and independent investigations suggested that Comcast interfered with over half of BitTorrent streams. After receiving a formal complaint about the practice, the Commission found “that Comcast's conduct poses a substantial threat to both the open character and efficient operation of the internet, and is not reasonable,” and ordered Comcast to cease the interference. However, the D.C. Circuit vacated the Commission's order in

Comcast.

94.

Madison River

and

Comcast-BitTorrent

—the anecdotes most frequently cited in favor of Title II regulation—demonstrate that any problematic conduct was quite rare. The more recent incidents discussed in the

Title II Order

also show that since 2008, few tangible threats to the

This text is long and has been trimmed here. Open the source document for the complete record.