Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
REGULATORY ISSUES As computer-based information technology has been adapted to an ever-widening array of economic and social pursuits, it has increasingly come within the purview of government regulation. This has occurred in part because many activi- ties the technology supports are themselves regulated. Electronic funds transfer systems, for example, are being noticeably shaped by the statutes and regulations govern- ing services that depository institutions and credit grantors provide to individuals. Equally important is the interaction between technological innovation and informa- tion policy generallyâthat is, between the processes by which technological advances are transformed into commer- cially available products and services and the bodies of law and interpretation that shape, directly, the way indi- viduals and organizations in our society assemble, dissemi- nate, and use information. The United States has never had a coherent national policy for the development and use of information tech- nology. Such information policy as we have had has been closely tied to the evolution of particular technologies, each conceptually distinct from the others, and each with its own special uses, markets, and associated inter- play of political and economic interests. As a result, the mail, the telegraph, the telephone, print publishing, broadcasting, cable television, automated data-processing, and satellite communication, each in its own time and each in its own way, have given rise to a separate body of law and interpretation and in many cases to a separate struc- ture of interpretative institutions. Each technology, more- over, has been approached from a somewhat different policy perspective. The conditions encouraging entry into the marketplace have varied greatly from one technology to another. Government approval has been required in some 26
27 instances, while free entry, subject only to antitrust con- straints, has elsewhere been the norm. Accordingly, com- petition has been forbidden, permitted, or required. First Amendment rights, privacy protection, copyright, and free- dom of information requirements have all been affirmed differently in different contexts. This pervasive fragmentation of policy initiative and responsibility could be sustained so long as there were valid technological grounds for distinguishing a letter from a phone conversation, a television image from a wired birthday greeting, and a telephone switchboard from a com- puter. Today, however, computer-based information tech- nology is calling all such distinctions into question and, with them, many of the premises, perspectives, and expec- tations on which our nation's information technology pol- icy, and particularly its regulatory policy, have heretofore been based. DATA PROCESSING OR TELECOMMUNICATIONS: THE FCC DEBATE The decade-long effort by the Federal Communications Com- mission to define a boundary between data processing and telecommunications provides a classic illustration of the regulatory challenges posed by computer-based information technology. As early as 1966, the Commission became aware that computing technology would play an increasingly cen- tral role in the operation of telephone and telegraph services, and that as a consequence many regulated common carriers would develop both a capacity and a desire to enter the growing market for commercially provided data- processing services. As the FCC repeatedly observed during its first formal inquiry into the subject: ... (C)ommon carriers "as part of the natural evolu- tion of the developing communications art" were rapidly becoming equipped to enter into the data processing field, if not by design, by the fact that computers utilized for the provision of conventional communi- cation services could be programmed additionally to perform data processing services.1 This technological convergence, moreover, was not occurring only within the telecommunications industry. Just as data-processing capability was becoming a basic tool of the telephone and telegraph industry, so also sig- nal transmission capability was becoming increasingly
28 indispensable to data-processing operations. As the FCC noted at the conclusion of its "Computer Inquiry I": ... (D)ata-processing cannot survive, much less develop further, except through reliance upon and use of communication facilities and services. To the Commission and the industries involved, it was apparent that this blurring of technological boundaries could raise some difficult policy issues. At the time, however, the issue that stood out was whether the Communi- cations Act of 1934 should be interpreted as requiring the FCC to regulate firms that provide communications-linked data-processing services on a commercial basis. The Communications Act of 1934 constitutes a Congres- sional determination that the public interest requires government regulation of any common-carrier communications activity. That determination was long interpreted by the FCC as requiring it to protect interstate providers of telephone and telegraph services from unregulated competi- tion. The rationale for such protection has been that without it the regulated carriers would be unable to gen- erate the revenues necessary to keep high-quality telephone and telegraph services continuously available to the public at affordable rates. Thus, in attempting to decide how it would treat communications-linked data-processing, the FCC has had to deal with two intimately related (as well as highly charged) regulatory questions: should communications-linked data-processing services be defined as a common-carrier communications activity and therefore subject to the same general rules as telephone and tele- graph services? should competition between data-processing firms and the established common carriers be encouraged, forbidden, or controlled? The first question, it so happened, sparked the con- troversy, since it was apparent to the FCC that both data processors and carriers were increasingly using combina- tions of computing and telecommunications technology to perform information-handling tasks that, from a techno- logical point of view, seemed virtually indistinguishable. The second question was equally important because no mat- ter how the Commission classified communications-linked data processing (i.e., as regulatable or not regulatable) the success or failure of the firms offering it as a commercial service might hinge on whether the established carriers could compete with them. For example, if the Commission were to decide that all or part of the
29 data-processing industry fell within its regulatory juris- diction, few other than the established carriers might have the financial resources necessary to obtain prior FCC approval of their service offerings. Conversely, if the Commission were to decide not to regulate data processing, it would, in effect, be releasing the carriers to enter the data-processing business without Commission supervision, and in doing so perhaps tempting them to use their pro- tected communications monopolies to secure formidable mar- ket advantage. The carriers, for instance, might use computer facilities or even revenues from their telephone and telegraph monopolies to subsidize, and thus artifici- ally price, their commercial data-processing offerings. Or they might provide their data-processing competitors with inferior communications equipment or services, or re- fuse to provide service on the grounds that to do so would be inconsistent with the intent of FCC policy.1 * At first glance, formally separating the carrier's regulated business enterprises from their unregulated ones might seem like a painless way to alleviate much of the concern about anticompetitive practices, and, in 1970, the FCC did adopt a regulatory policy based on a concept of "maximum separation." The carriers were precluded from otherwise disposing "of any capacity on computer systems utilized...for the provision of common carrier communica- tions services," and, in addition, the Commission stipu- lated that they might enter the commercial data-processing market only through separately incorporated affiliates.5 Nonetheless, as the Commission recognized, this was only a partial solution to a partially articulated policy dilemma. In principle "maximum separation" did little more than require the carriers to maintain two sets of accounts, one for their regulated communications services and one for their independent, and thus unregulated, data-processing services. It did not provide anyone with the ability to say whether the communications or data-processing operations thus accounted for were the ones intended. Yet, unless the Commission could delineate what was intrinsically communi- cations and what was not, unregulated data-processing ser- vices would always be vulnerable to competition from the regulated carriers' technologically similar communications ones. Likewise, in order to fulfill its obligation to pro- tect the carriers from unsupervised competition in the telecommunications domain, the Commission would inevitably be called upon to decide whether services marketed as com- mercial data processing were in fact communications ser- vices in disguise and, thus, properly subject to regulation.
30 As an illustration of how difficult it can be to draw such distinctions, consider the case of message switching. Message switching is a form of information transmission in which signals, instead of going directly from the sender to the receiver, are passed through one or more routing facilities that determine the path they should take and the priority they should have in the queue at any given routing point. Historically, message switching has been the basic common carriage service provided by the telegraph industry. Today, however, it is also used in data-processing operations where the workload generated by one or more re- mote terminals is shared among several computers. To com- plicate matters further, the telegraph industry is itself in the course of shifting from conventional to computer- based message switching. Under such circumstances, how might the FCC decide, on technological grounds, that mes- sage switching in the telegraph context constitutes regu- latable telecommunications, while elsewhere it may be properly classified as unregulated data-processing? The method by which the FCC chose to extricate itself from such ambiguities was to begin defining services as either communication or data processing depending on the degree to which each exploits the capabilities of the computer. The Commission defined message switching as a communications service on the grounds that, in contrast to data processing, it does not involve any alteration of the information being transmitted. From the Commis- sion's perspective, message switching was pure informa- tion transmission. This could mean, however, that any data-processing service that employed message switching could be regulated as a communications service, and, as indicated, the Commission, for a variety of reasons, was reluctant to reach that conclusion. Thus, the Commission distinguished a special class of "hybrid" services that it further divided into hybrid communications services (in which data processing plays only a minor role), and hybrid data-processing services (in which communications plays the minor role). Which was which, the Commission said, it would decide for itself after evaluating "the primary thrust of the service offered."6 THE WEAKNESSES OF THE HYBRID SERVICES APPROACH It can be difficult to describe policy making processes and outcomes of this sort without appearing to criticize the regulatory agency responsible. The FCC, for its part,
31 was doing no more than interpreting general rules and ob- jectives laid down in its statutory mandate. That mandate, in effect, enjoined it to differentiate among service of- ferings that historically had been distinguishable by the capabilities of the technologies they employed, while, as a practical matter, the direction of technological change was making distinctions among those capabilities harder and harder to draw. The problem in other words was not that the path the Commission chose was unreasonable, given the circumstances, but rather that it was beholden to an increasingly anachronistic conception of its regulatory functions. By adopting the hybrid services concept, moreover, the Commission inadvertently inhibited some of the changes it had hoped to find a way of accommodating, since, as subse- quent events have shown, one of the chief effects of the hybrid services approach has been to discourage innovation in the hybrid services area. The Commission's action created a situation in which a firm wishing to market a communications-linked data-processing service typically had two choices: to request prior FCC evaluation with all the delay and expense that could entail, or, alternatively, to venture into the unregulated marketplace and wait to see if anyone challenged its right to be there. The conse- quences of posing such a dilemma for potential service pro- viders have never been systematically assessed, but there seems to be little doubt that they impeded the introduction of new services made possible in principle by cost effi- ciencies inherent in new generations of computers. One clear result is that much needed productivity gains are being foregone in areas ranging from manufacturing and banking to postal service and education, and the loss is being felt not only domestically, but also when U.S. firms offer their products and services for sale abroad. Sensing this, and also being an early convert to the "regulatory reform" movement of the mid-1970's, the FCC, in 1976, launched a second inquiry into the technical and func- tional relationships between data processing and communica- tions. Known as "Computer Inquiry II," the investigation has resulted in a recent Commission decision to abandon the hybrid services approach in favor of one that distinguishes between "basic transmission services," which the FCC will continue to regulate, and "enhanced services," which it proposes not to regulate.^ In its Final Decision announcing this new policy, the Commission defines a "basic transmission service" as
32 the offering of transmission capacity between two or more points suitable for a user's transmission needs and subject only to the technical parameters of fidelity or distortion criteria or other condi- tioning. Use internal to the carrier's facility of computing techniques, bandwidth compression techniques, circuit switching, message or packet switching, error control techniques, etc. that facilitate economical, reliable movement of infor- mation does not alter the nature of the basic service. Computer memory or storage, for example, may be used within the network to facilitate the transmission of information from origination to destination, provided that "the car- rier's basic transmission network is not used as an infor- mation storage system."10 A "basic transmission service," in short, is one in which the communications carrier essentially offers a pure transmission capability over a communications path that is virtually trans- parent in terms of its interaction with customer supplied information.11 An "enhanced service," by contrast, is defined as one that uses the telecommunications network to accomplish something more than pure transmission.12 In the typical enhanced service, computer processing applications are used to act on the content, code, protocol, and other aspects of the subscriber's information. In these services additional, different, or restructured information may be provided the subscriber through various processing applications performed on the trans- mitted information, or other actions can be taken by either the vendor or the subscriber based on the content of the information transmitted.13 For a service to qualify as "enhanced," information need not be changed by either the vendor or the subscriber, since the enhanced category would encompass an "electronic mailbox" service that "simply involves subscriber interac- tion with stored information."11* Typically, however, an enhanced service would involve considerably more than sim- ple storage and retrieval capability. The Commission of- fers as its principal illustration of such a service a time-sharing network in which
33 computer facilities are structured in a manner such that the customer or vendor can write its own cus- tomized programs and, in effect, use the... network for a variety of electronic message service operations.15 Key to the practicability of this twofold categorization of carrier services is the Commission's corollary decision to reduce the scope of its 1970 ruling on "maximum separa- tion." As explained earlier, the FCC, at the close of Com- puter Inquiry I, required a regulated communications carrier wishing to enter the unregulated data processing market to do so through a separately incorporated subsidiary. Each such subsidiary was required to maintain its own books, to have its own officers and operating personnel, and to use computer equipment and facilities separate from those used in providing regulated communications services.15 This degree of organizational separation was seen as a necessary safeguard if carriers were to be prevented from using reve- nues or other resources associated with their regulated communications operations to secure competitive advantages for their unregulated data processing ones. Now, however, the FCC has concluded that only carriers "under direct or common control of AT&T17 or GTE18 have the capacity to engage in such anticompetitive behavior."19 In the Com- mission's opinion, only those corporations possess the requisite combination of financial resources and dominant market position. Thus, only AT&T and GTE will be subject to maximum separation rules. All other carriers will be allowed to offer whatever enhanced services they wish as direct adjuncts to their basic service operations. This narrowing of the scope of the maximum separation requirement is thought to be crucial in at least two re- spects. First, by easing the requirements for entry into the enhanced services market, more, better, and more varied services are expected to become available to users. Second, the FCC will no longer have to decide on a case-by-case basis whether each new service offered by a communications carrier constitutes communications or computing, or, to use the terminology of the final decision, whether it consti- tutes a "basic transmission" service or an "enhanced" one. The FCC will not be wholly freed of the need to make such distinctions, since questions concerning the proper cate- gorization of certain AT&T and GTE offerings may still arise. The Commission has noted that its definition of an enhanced service effectively precludes the use of code r\ n and protocol conversion as part of a basic transmission
34 service, and that as a result, carriers subject to the maximum separation requirement may be deprived of oppor- tunities to increase the utility of their communications channels.21 If that proves true, one may expect the af- fected carriers to appeal to the Commission for relief. Nonetheless, the FCC appears to have found a way to sim- plify greatly its regulatory task. Indeed, it appears that the key issue for the Commission now is not whether it will be able to develop new regulatory policy suited to the growing convergence between computing and telecommuni- cations technology, but rather whether it will be able to carry out that policy. There is a possibility, for example, that aspects of the FCC's decision will be challenged on the grounds that they exceed the Commission's regulatory authority or, con- versely, constitute an illegal exercise of discretion not to use that authority. One point of potential controversy is whether the Communications Act of 1934 allows the FCC to impose varying degrees of regulation on carriers provid- ing the same type of serviceâi.e., whether the maximum separation requirement can be applied selectively. An- other is whether the Commission is statutorily required to regulate enhanced servicesâi.e., whether it has the author- ity to choose not to do so. Questions have also been raised by the U.S. Department of Justice as to the effect of the FCC's new policy on a 1956 consent decree that for- bids AT&T and its affiliates to engage in any commercial activity not regulated by the Commission,2I* and, in addi- tion, there are numerous complications stemming from the FCC's decision on basic vs. enhanced services that is closely linked to an accompanying rulemaking on the sale and installation of so-called "customer premises equipment (CPE)."25 The latter poses particularly difficult problems since its execution requires a major restructuring of the rates charged for interstate telephone service as well as extensive cooperation between federal and state regulators. Many of these hurdles may be overcome by timely Con- gressional action on proposed amendments to the Communica- tions Act, while others may be settled through litigation. The point to be emphasized is that the issues addressed by the FCC's latest action are still far from settled, even though the Commission has been working to resolve them for more than 13 years. Furthermore, one can expect that in the future the Commission will be confronted with regulatory questions of even greater complexity as computer- based information technology becomes the essential under- pinning of communications modalities such as broadcasting
35 and cable television. As the technology that is economi- cally advantageous to each of these becomes more and more the same, and as new kinds of hybrid services become in- creasingly feasible and attractive, the inadequacy of con- ventional approaches to establishing and implementing regulatory objectives will be dramatically, and often pain- fully, revealed. Efforts to rethink those objectives and to devise more predictable strategies for implementing them are currently afoot in the Congress. At the moment, however, primary attention is focused on modifying the internal structure of the several communications industries and on the future role of the FCC. Legislation recently before the Congress would replace the Commission with a weaker regulatory agency whose principal mandate would be to stimulate fur- ther competition within the telecommunications industry, while looking to the Executive branch for broad policy guidance. Whether such legislation will ever be enacted is currently hard to say, but however the matter is de- cided the effects will surely be felt in many areas. Today, for example, there is confusion whether elec- tronic funds transfer and electronic message service will be under FCC regulatory jurisdiction or will be assigned to or shared with other agencies, such as the bank regula- tory authorities and the U.S. Postal Service. Making such collateral assignments of regulatory responsibility may prove to be a practical alternative. If technology no longer provides a reliable basis for bounding regulatory jurisdictions, a regulatory structure organized around primary application areas may be the sensible alternative. The difficulty is that today judgments of that type can be made more on the basis of habit or expediency than on the basis of carefully considered, independent evaluations of their implications. THE ROLE FOR INDEPENDENT POLICY RESEARCH Coping with the many regulatory issues that computer-based information technology raises is clearly a task for govern- mental decision making processes that give full voice and consideration to the interests of the potentially affected parties. Fact gathering and analysis outside of those pro- cesses can only play a supporting role. That role, how- ever, need not be a small one, nor need it be confined to purely technical issues.
36 Although a technical consensus on matters such as defi- nitions, feasibility, and costs is sorely needed, public policy decisions, or nondecisions, must also be based on a much heightened understanding of market peculiarities, financing conditions, and the propensity of public and private sector users to exploit the technology's capabili- ties. Just as each of these elements can expand or limit the future flow of resources to a service or market, so also can they expand or limit the variety of policy options to be considered. In addition, because the pressure for timely government action is likely to grow rather than diminish, there will be corresponding pressure to keep the decision making pro- cess focused on issues that have already been identified. This, too, will tend to prevent questions that ought to be asked, and options that ought to be weighed, from being given the careful consideration they deserve. In short, a fresh look that takes full advantage of previous efforts, but also ventures well beyond the cur- rent preoccupation with new or revised roles for exist- ing institutions, seems to be very much in order. As a first step, technical, economic, legal, and sociological expertise should be jointly applied to the task of design- ing a program of independent research and analysis to sup- port the public decision making process. Developing a capacity to articulate the implications of alternative policy proposals and to highlight and examine implicit policy assumptions should be particularly stressed. For example, recent history seems to suggest that regu- latory policy should have as its central objectives (1) giving full scope to innovation and (2) assuring that the public interest is adequately protected. Yet, because those objectives are not always compatible, and because there are more or less optimal ways to make and implement the choices and adjustments they demand, their interaction needs to be explored carefully within the context of alter- native regulatory strategies. Similarly, although many now seem to believe that regu- lated and unregulated markets can be effectively interwoven under circumstances of diminishing government supervision, the circumstances under which that assumption may or may not be valid need to be more thoroughly considered. Is it reasonable, for example, to believe that cross subsi- f\ ~i dization and other predatory practices can be eliminated if a corporate family is still allowed to provide both regulated and unregulated services? Has the point been reached at which conscious policy decisions can create
37 highly competitive service markets, with little or no direct government involvement? Are there some counter- vailing economic or technological trends that, in the absence of government supervision, will tend to promote increasing concentration of control over the resources for innovation? The prospect of interaction between regulatory objec- tives and other types of policy objectives might also be examined. If stimulating innovation is to be a central policy objective, are there, for instance, aspects of fed- eral and state tax policy that will tend to keep the com- petitive marketplace from providing that stimulus? Are there dimensions of antitrust policy that need to be re- thought in the light of their observable or projected im- pact on incentives to innovate? Thus, some contend that antitrust constraints now tend to keep the fruits of basic research from reaching the marketplace as rapidly as they should, while others maintain that removing those con- straints would allow basic research knowledge to be used in predatory ways. At this juncture, it is possible to do no more than illustrate the kinds of issues and questions that indepen- dent policy analysis might address. These should indicate, at the least, the scope of the program that is envisaged and also the importance of starting from a strong, multi- disciplinary perspective.
38 NOTES AND REFERENCES 1. Regulatory & Policy Problems Presented by the Interde- pendence of Computer & Communications Services and Facilities, 28 FCC 2d 267 (1971) (Final Decision), 36 Fed. Reg. 5346. 2. Ibid. 3. All the carriers except AT&T, which, by the terms of a 1956 consent decree, has been prohibited from engaging in any commercial activity not regulated by the FCC. 4. For a thorough and lucid analysis of the many issues before the FCC in this area, see Paul J. Herman, "Computer or Communications? Allocation of Functions and the Role of the Federal Communications Commission," in Anthony G. Oettinger, Paul J. Herman, and William H. Read, High and Low Politics: Information Resources for the '80s. Cambridge, Mass.: Ballinger Publish- ing Company, 1977. 5. FCC Final Decision, op. cit. 36 Fed. Reg. 5349. This ruling did not apply to AT&T, which by the terms of a 1956 Consent Decree has been permitted to engage only in those activities that involve the offering of a communications service or facility and are subject to public regulation. 6. Regulatory and Policy Problems Presented by the Inter- dependence of Computer and Communications Services and Facilities, 28 FCC 2d 291 (1970) (Tentative Decision), 35 Fed. Reg. 5822. For a more detailed analysis see Oettinger, et al., op. cit., p. 157-159. 7. Amendment of Section 64.702 of the Commission's Rules and Regulations, Notice of Inquiry and Proposed Rule- Making, 64 FCC 2d 103 (1976), 41 Fed. Reg. 33563; Sup- plemental Notice of Inquiry and Enlargement of Proposed Rulemaking, 64 FCC 2d 771 (1971), 42 Fed. Reg. 13029. 8. In the Matter of Amendment of Section 64.702 of the Commission's Rules and Regulations (Second Computer Inquiry), Docket No. 20828 (Final Decision), 45 Fed. Reg. 31319.
39 9. Ibid., pp. 31333-31334. 10. Ibid., p. 31334. 11. Ibid. 12. Ibid. 13. JJbid. 14. Ibid. 15. Ibid. 16. Ibid., p. 31349. 17. I.e., The American Telephone & Telegraph Corporation. 18. I.e., the General Telephone & Electronics Corporation. 19. In the Matter of Amendment of Section 64.702 (Final Decision op. cit., p. 31321. 20. As the Commission explains in a footnote: In this context, "code" means the binary represen- tation of alphanumeric and control characters. Thus an enhanced service may modify the transmitted bit stream to change it from the ASCII code to the EBCDIC code, which a basic service may not. "Pro- tocols" govern the methods used for packaging the transmitted data in quanta, the rules for control- ling the flow of information, and the format of headers and trailers surrounding the transmitted information and of separate control messages. Ibid., p. 31334. 21. Ibid. 22. Ibid., p. 31327. 23. Ibid., p. 31329. 24. Ibid., pp. 31362-31363. 25. Ibid., pp. 31340-31344. 26. Ibid., pp. 31345-31346.
40 27. Cross subsidization is the practice of using the reve- nues or facilities of one enterprise to help pay the cost of offering the products or services of another enterprise for sale at less than their full market value. In the FCC context, cross subsidization would typically involve the transfer of such resources from a regulated monopoly to an unregulated business enter- prise run by the monopoly or one of its subsidiaries.