Federal Internet Law & Policy
An Educational Project
Computer III Dont be a FOOL; The Law is Not DIY

The Setting

This time the Commission managed to have four years pass without reopening its proceeding. It is 1985. The Domain Name System had just been introduced the previous year. The first commercial ISP had yet to be established. The National Science Foundation Network ("NSFNET") would come online in 1986.

In 1984, AT&T had gone through divestiture forming the BOCs. The BOCs for their part received official blessing from Judge Greene, permitting them to begin to enter the enhanced services market in 1987 and to fully enter the market in 1991. [ONA Review ¶ 29] [Accounting Safeguards ¶ 3 n 5 (“The information services restriction was modified in 1987 to allow BOCs to provide voice messaging services and to transmit information services generated by others. In 1991, the restriction on BOC ownership of content-based information services was lifted.”)] In 1985, the Commission launched Computer III, [CIII NPRM] initiating the last phase of the regulatory proceeding that would have so many implications for the deployment of the Internet before the commercial Internet truly broke.

The Issue

The conceptual framework of the Computer Inquiries had been settled in Computer II with the establishment of the basic versus enhanced service dichotomy. The policy objectives remain the same. What changed was implementation.

Driving Computer III was the perception that the separate subsidiary requirements of Computer II "impose significant costs on the public in decreased efficiency and innovation that substantially outweight [sic] their benefits in limiting the ability of AT&T and the BOCS to make unfair use of their regulated operations for the benefit of their unregulated, enhanced services activities." [CIII R&O 1996 ¶ 3]

In Computer III, after reexamining the telecommunications marketplace and the effects of structural separation during the six years since Computer II, the Commission determined that the costs of structural separation out-weighed the benefits, and that nonstructural safeguards could protect competitive ESPs from improper cost allocation and discrimination by the BOCs while avoiding the inefficiencies associated with structural separation.

[CIII Order 1999 ¶ 7] Believing that it could achieve appropriate safeguards, perhaps develop a new and more progressive framework, and also eliminate the cost of the separate subsidiary, the Commission sought to migrate from structural safeguards to non-structural safeguards. In order to do so, the Commission had to establish a scheme that would satisfy its original concerns regarding anticompetitive behavior and continue to make available an open communications platform. The Commission's solution was, in the short term, Comparatively Efficient Interconnection ("CEI") and, in the long term, Open Network Architecture ("ONA").

The Resolution

Comparatively Efficient Interconnection

CEI was seen as an interim solution while BOCs create ONA plans. Under CEI, BOCs would be permitted to enter into enhanced services markets on a non-structurally separated basis (a separate subsidiary was no longer needed; the ESP could be integrated into the BOC) on the condition that they make available [CIII Order 1999 ¶ 4, 11-12 (FCC rules permitted the posting of CEI plans to the company website)] [CIII Order on Recon 1999 ¶ 6] CEI plans. These CEI plans were intended to detail what the BOC was provisioning to its affiliated ESP; BOCs would be required to make these provisions available to all other non-affiliated ESPs on the same terms and conditions.150

A CEI plan must include information on interface functionality, unbundling of basic services, resale, technical characteristics, installation, maintenance and repair, end-user access, CEI availability, minimization of transport costs, and recipients of CEI. [CIII Further 1998 ¶ 4] [BOC's Joint Petition ¶ 3] [Ameritech’s CEI Plan ¶ 16 (“The CEI requirements are designed to give ESPs equal and efficient access to the basic services that the BOCs use to provide their own enhanced services.”)].

Open Network Architecture

"ONA is the overall design of a carrier's basic network services to permit all users of the basic network, including the information services operations of the carrier and its competitors, to interconnect to specific basic network functions and interfaces on an unbundled and equal-access basis. The BOCs and GTE through ONA must unbundle key components, or elements, of their basic services and make them available under tariff, regardless of whether their information services operations utilize the unbundled components. Such unbundling ensures that competitors of the carrier's information services operations can develop information services that utilize the carrier's network on an economical and efficient basis." [CIII Order 1999 ¶ 8 n 17]

The next step was the move toward Open Network Architecture. This was a radical approach to the issue. While the Commission was in retreat on the notion and value of separate subsidiaries, ONA imposed a progressive vision of the network. ONA required BOCs to break their networks down into basic building blocks, and to make those building blocks available to ESPs to build new services. [CIII Remand 1995 ¶ 15-16] Although not identical to unbundled network elements, the BOCs would be required to break apart the basic service offering for the benefit of the ESP market. These BOC building blocks would be divided among

[CIII Further 1998 ¶ 26] [ONA Review 1988 ¶ 56]

BOCs were required to file ONA plans regardless of whether they entered the ESP market. [Ameritech’s CEI Plan ¶ 7 n 18] [BOC’s Joint Petition ¶ 26] Having successfully filed an ONA plan with the Commission, BOCs would then be permitted to provide integrated ESP services without filing a CEI plan. [BOC’s Joint Petition ¶ 3] There was also guidance on how ESPs could request the provision of new services. [CIII Further 1998 ¶ 81-84]

The Computer III proceeding also set forth a set of other safeguards that fell upon different entities:


Things did not go quite as planned. In California III, the Ninth Circuit reviewed the Commission's move from structural to non-structural safeguards and:

found that, in granting full structural relief based on the BOC ONA plans, the Commission had not adequately explained its apparent "retreat" from requiring "fundamental unbundling" of BOC networks as a component of ONA and a condition for lifting structural separation. The court was therefore concerned that ONA unbundling, as implemented, failed to prevent the BOCs from engaging in discrimination against competing ESPs in providing access to basic services.162

[CIII Further 1998 ¶ 15] [California v FCC 1994] The Court therefore vacated and remanded the proceeding back to the FCC.

On remand, the Commission concluded that the court in California III vacated only the Commission's ONA rules, not the CEI rules. Therefore, the Commission issued the Interim Waiver Order that permitted BOCs to provide enhanced services if they complied with the CEI rules. [CIII Further 1998 ¶ 16] In addition, BOCs must comply with procedures set forth in the ONA plans that they had already filed with and had been approved by the Commission. [Ameritech’s CEI Plan ¶ 6] [BOC’s Joint Petition ¶ 22] [CIII Remand 1995 ¶ 9-12] [Bell Atlantic CEI 95 ¶ 4] The Commission also released a Further Notice of Proposed Rulemaking in order to resolve the issues raised in California III. [CIII Remand 1995] In 2002, the Commission released the Broadband Notice of Proposed Rulemaking, which subsumed the Computer III proceeding and is currently pending.


The Commercial Internet eXchange ("CIX") objected to the movement to non-structural safeguards, arguing that this created a problem with enforcement. Recognizing validity to the CIX objection, the Commission stated:

We believe that competitive ISPs will themselves monitor CEI compliance vigilantly, and will call the Commission's attention to any failure by a BOC to follow through on its CEI responsibilities. . . . The Commission will not hesitate to use its enforcement authority, including the Accelerated Docket or revised complaint procedures, to review and adjudicate allegations that a BOC is falling short of fulfilling any of its CEI obligations.

[CIII Order 1999 ¶ 15] Note, however, that this does create certain structural oddities. First, an unregulated industry, with little knowledge of the FCC, is asked to watch a regulated industry. Second, small companies are asked to watch the largest corporations in the United States. Third, ISPs are placed in a position of filing complaints against their sole supplier of a crucial facility. Fourth, contrary to normal jurisprudence, the party that lacks the information has the burden of moving (normally, all things being equal, the party with the information has the burden of moving-in this case, the burden is on the ISPs, because the information is held by the BOCs).

Legacy of Computer III

The legacy of Computer III was first and foremost an affirmation of the policy goals set forth in Computer I and Computer II. Computer III did not alter the fundamental philosophy of the Computer Inquiries. Concern for anticompetitive behavior and maintaining an open communications platform was retained. Likewise, the conceptual framework of the basic versus enhanced dichotomy that tracks the technical layers of the network was affirmed.

Computer III altered implementation by initiating new, novel, and even progressive experiments in opening up the communications bottleneck for the benefit of enhanced services. This marked a migration away from structural safeguards. It also signified renewed emphasis on opening up the facilities network for the benefit of the enhanced service or computer network operating on top of the physical network. It affirmed that while enhanced services remain unregulated, the Computer Inquiry policy is designed for the direct benefit of those enhanced services.

The transition from Computer I to Computer II marked a radical evolution of the conceptual framework, with implementation remaining fundamentally the same. The transition from Computer II to Computer III marked a radical evolution in the implementation, with the conceptual framework remaining the same.

© Cybertelecom ::