The legacy of the Federal Communications Commission's computer inquiries.

• Author: Cannon, Robert

1. INTRODUCTION II. COMPUTER I (1966) A. The Setting 1. A Better Mouse Trap 2. Western Union 3. Big Iron and New Networks B. The Issue C. The Policy 1. Classification 2. Regulation 3. Safeguard: Maximum Separation D. Legacy of Computer I III. COMPUTER II (1976) A. The Setting B. The Issue C. The Resolution 1. Basic versus Enhanced Service Dichotomy a. Basic Services b. Enhanced Services c. Adjunct Services d. Protocol Processing e. The Telecommunications Act of 1996 and "Information Services" 2. Safeguards a. Maximum Separation to Structural Separation b. Unbundling D. Layers E. Legacy of Computer II IV. COMPUTER III (1985) A. The Setting B. The Issue C. The Resolution 1. Comparatively Efficient Interconnection 2. Open Network Architecture 3. Litigation 4. Enforcement D. Legacy of Computer III V. CONCLUSION I. INTRODUCTION

   In the 1960s, the Federal Communications Commission ("FCC" or "Commission") awoke to the reality of powerful computers running communications networks, and communications networks over which humans interacted with really powerful computers. Computer services were a disruptive technology. They were substitute services for traditional incumbent communication services. They were highly competitive, highly innovative, and had low barriers to entry. They showed every promise of playing a vital role in the United States economy. In addition, these computer network services were dependent upon the underlying communications network. Thus, the unregulated computer services were simultaneously substitute services for the traditional regulated communications networks and also dependent upon them.

   Meanwhile, the communication network services were using gigantic mainframe computers ("big iron") to run their networks. During network peaks, mainframe computers were preoccupied with operating the networks. During off-peaks, these computers had excess capacity. The telephone companies knew a good thing when they saw it and wanted to get into the computer services market, taking advantage, in part, of their inexpensive excess off-peak mainframe capacity. Thus, the telephone companies became simultaneously the supplier of the crucial transmission capacity and a competitor in the computer services market.

   The FCC has struggled with the regulatory treatment of computer networks over communications networks ever since. In 1986, the Commission stated:

   The regulatory issues spawned by the technical confluence of regulated communications services and unregulated [computer networks] have been among the most important matters this Commission has dealt with over the past 20 years. Indeed, during this period, we have addressed these issues, in one proceeding or another, on a virtually continuous basis, as we have sought to revise and refine our regulatory approach in light of rapidly changing technological and marketplace developments. (1) The history of the FCC and the computer networks, particularly the Internet, is now thirty-five years old. To say that the FCC does not regulate the Internet is to miss the lessons of this history. While it is true that computer networks are unregulated, computer networks were very much a part of the Commission's policy. They were the intended direct beneficiaries of the Computer Inquiries. Safeguards were imposed on common carriers for the benefit of computer networks. In addition, this is not a history of technologically biased regulation, segregating one computer from another based on the technology employed. Rather, this is a market policy, segregating competitive markets from noncompetitive markets. Finally, the conceptual framework follows a Layered Model of Regulation. The separate layers permitted, even created, separate markets (i.e., telephone service, Internet service, application service, and content). These separate markets created separate regulatory policy.

   The Computer Inquiries have been referred to by some as "wildly successful." (2) They were a necessary precondition for the success of the Internet.

2. COMPUTER I (1966)

   1. The Setting

      1. A Better Mouse Trap

         It is the 1960s. The FCC faces a problem. At this time there exists a communications network that offers basic communications service. This communications network is provided by the incumbent monopoly, Ma Bell, also known as AT&T. This network has been traditionally regulated by the FCC. It was built in a regulatory environment as a sanctioned monopoly with ratepayer fees. (3)

         The problem was that someone figured out how to build a better mouse trap. Someone figured out how to use computers with this network. Someone figured out how to enhance the network by adding devices at the ends of the network, layering protocols on top of the network, and achieving data processing using remote terminals. Ultimately these innovations would evolve into computer networks. These enhancements were dependent upon the underlying communications monopoly and came with the marvelous promise of economic expansion, innovation, and competition. These enhancements, however, also threatened to be a substitute for regulated services, and regulated services threatened to be a bottleneck in the way of the growth of these services.

      2. Western Union

         In the 1960s, the FCC's common carrier authority covered not merely Ma Bell; Western Union also fell under Title II of the Communications Act of 1934. The Western Union telegram service was, in retrospect, an interesting service. The service took an order for a message from a user. The user provided a destination address and content for the message. The message then was inserted into the Western Union system. However, if the message originated in Baltimore for a destination in Los Angeles, it did not go straight through a wire from Baltimore to Los Angeles. Rather, the operator in Baltimore, not knowing the full path of the transmission to Los Angeles, knew instead the next hop in the general direction of Los Angeles, and forwarded the message to that next hop. At the next hop, which might have been Chicago, an operator received the telegram, read only the header with the address, and then forwarded the message back into the network in the general direction of the destination. In this way, the message worked its way through the network, being stored and then forwarded, hopping from node to node in a best effort, until it reached its final destination and was stored until delivered to the recipient. (4)

         In the 1960s, it dawned on users of mainframe computers that they could take advantage of the excess capacity of the mainframes to send messages to each other. Alfred may log onto the mainframe from one remote terminal, and leave a message for Beth. Beth would then log onto the mainframe from another remote terminal, perhaps in another state, and receive the message. Eventually computer message-switching (5) became a commercial service that did not simply store messages on a mainframe, but transmitted messages through computer networks. Alfred would create a message and hit send. The message would go to the first e-mail server, which would read the address, and then send the message on to the next hop in the network. The next hop would read the address and act accordingly. When the message reached its destination computer, the message was stored until accessed by the recipient, Beth.

         These two things look very similar to each other. However, one was regulated; the other was not. One was expensive; the other one was cheap, and avoided regulatory fees. One is a substitute service for the other. (6)

         There are two things to be taken away from this. First, message-switching was dependent upon a regulated underlying telephone monopoly for transmission. Second, unregulated message-switching was a substitute service for the regulated telegram services of Western Union. (7) What exactly to do with message switching was one of the significant drivers of the Computer I inquiry. (8)

      3. Big Iron and New Networks

         This was a moment of major expansion of the American economy. Big iron mainframe computing had taken hold and was becoming big business. Mainframe computing was also evolving with the advent of time sharing and remote terminal access. (9) The role of IBM, computer manufacturers, and data processing services in the economy had grown and promised continued growth. (10) There were in-house computer services, computer service bureaus, and specialized computer services such as stock quotation services. (11) Computers were being used to facilitate President Kennedy's space race, advance the Cold War, run communications networks that replaced human operators, and re-create the way business was conducted. (12) The Internet would not be born until the end of the 1960s. (13) The United States government responded to the 1950s Soviet launch of Sputnik with, among other things, the establishment of the Advanced Research Projects Agency ("ARPA"). ARPA's computer research program, headed by individuals such as J.C.R. Licklider and Larry Roberts, led a team of researchers to develop the ARPANet. On October 25, 1969, ARPANet went online, transmitting its first message between computers at the University of California at Los Angeles and the Stanford Research Institute. Originally, the government-run ARPANet used the Network Control Protocol; it did not migrate to the Internet Protocol ("IP") for 14 years. (14)

   2. The Issue

      In the 1960s, the FCC faced a problem of something the Commission referred to as "convergence." (15) There were computers that facilitated the operation of the communications network and there were computers with which humans interacted. (16) What should the Commission make of these computers? How should they be treated and how do they fall within the regulatory scheme? What type of jurisdiction did the FCC have over these computers and should data processing services be regulated under Title II of the Communications Act of 1934? Should the FCC be concerned that some of those regulated...