National Academies Press: OpenBook

Broadband: Bringing Home the Bits (2002)

Chapter: 5 Broadband Policy and Regulation

« Previous: 4 Technology Options and Economic Factors
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 167
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 168
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 169
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 170
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 171
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 172
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 173
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 174
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 175
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 176
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 177
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 178
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 179
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 180
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 181
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 182
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 183
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 184
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 185
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 186
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 187
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 188
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 189
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 190
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 191
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 192
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 193
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 194
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 195
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 196
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 197
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 198
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 199
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 200
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 201
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 202
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 203
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 204
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 205
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 206
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 207
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 208
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 209
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 210
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 211
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 212
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 213
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 214
Suggested Citation:"5 Broadband Policy and Regulation." National Research Council. 2002. Broadband: Bringing Home the Bits. Washington, DC: The National Academies Press. doi: 10.17226/10235.
×
Page 215

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.

5 Broadband Policy and Regulation This chapter provides an introduction to the policy context surround- ing broadband deployment and discusses specific issues that have shaped the reasoning of the Committee on Broadband Last Mile Technology and that underlie its recommendations. A more detailed history of U.S. regu- lation related to broadband appears in Appendix B, which is recom- mended for any reader not familiar with the complex regulatory context within which broadband is being deployed. Note that this chapter does not contain the committee’s policy recommendations, although it lays part of the foundation for them. The committee’s findings and recommendations are presented in “Summary and Recommendations” at the beginning of this volume. THE CONTEXT FOR BROADBAND POLICY Broadband—as an extension or phase of the Internet—has been im- bued, in media coverage and popular debate, with revolutionary promise that cuts across traditional policy segmentation. Residential broadband has ushered in an era of considerable technological innovation and flux. At the same time that a diverse set of technologies, which are character- ized by different performance characteristics (bandwidth, symmetry, transparency, and so on), are available for reaching customers, a powerful convergent platform—the Internet and its core technologies—is increas- ingly favored for delivery of content, applications, and services. Enabled by this flexible, general-purpose delivery platform, a multiplicity of ap- plications and content supported through a variety of business models 167

168 BROADBAND has emerged. The combination of technological and associated business changes generates considerable uncertainty and questions about what regulatory or other approaches are suited to meeting desired goals in light of this uncertainty. Broadband has been targeted by traditional political players in the various policy arenas, and it has catalyzed the formation of new political alliances. It is subject to past policy developed in part for other telecom- munications technologies and is the focus of a number of efforts to shape new telecommunications policy. The previous chapter explained how the characteristics of the different technology options are just a piece of the broadband puzzle. Because the profitability and growth prospects of dif- ferent kinds of entities are affected by government decisions about what kind of entity can provide what kind of service, and when and where and how it can do so, the policy context has a significant impact—there are no pure investment decisions, and political activity aimed at shaping the context is rampant. While many speak of a desire to deregulate, the na- ture and terms of regulation have become part of the competitive process. Moreover, there is something fundamentally highly political in the nature of communications technologies and services, beginning with the impor- tance of communications media in the political process itself.1 The recent introduction of a number of pieces of legislation aimed at promoting broadband is another indicator of heightened interest and sometimes in- tense politicization. Proposed measures include tax credits, grants, subsi- dized loans, and other financial incentives for deployment in underserved or rural areas; support for research on broadband technologies for rural areas; grants for community planning efforts; changes in the regulation of incumbent local exchange carriers; and changes in universal service fund rules.2 Viewed through the lens of telecommunications policy, broadband involves a system with players and rules at federal, state, and local levels and a long history of political activity that features industry associations old and new, consumer- and issue-advocacy organizations (and consider- 1For example, the original schemes for allocating radio and television licenses had a political connection, with licenses allocated geographically. 2Bills that would provide financial incentives include H.R. 267, Broadband Internet Ac- cess Act of 2001; H.R. 1415, Technology Bond Initiative; H.R. 1416, Broadband Expansion Grant Initiative; H.R. 1697, Broadband Competition and Incentives Act; H.R. 2139, Rural America Broadband Deployment Act; H.R. 2401, Rural America Digital Accessibility Act; H.R. 2597, Broadband Deployment and Telework Incentive Act; H.R. 2669, Rural Telecom- munications Enhancement Act; S. 88, Broadband Internet Access Act; S. 150 Broadband Deployment Act; S. 426, Technology Bond Initiative; S. 428, Broadband Expansion Grant

BROADBAND POLICY AND REGULATION 169 able activity by lawyers and lobbyists for all parties) seeking to influence legislation, administrative rule making, and court decisions. Organiza- tions addressing broadband as part of their lobbying activity have prolif- erated; they range from mainstream telecommunications trade associa- tions (such as the United States Telecom Association, the National Cable and Telecommunications Association, and the Organization for the Pro- motion and Advancement of Small Telephone Companies) to associations of new telecommunications competitors (e.g., the Association for Local Telecommunications Services and the Competitive Telecommunications Association) to technology-specific associations (e.g., DSL Forum and the Home Phoneline Networking Alliance), broadband issues-focused asso- ciations (such as the Competitive Broadband Coalition and the OpenNET Coalition), and consumer advocacy organizations (such as Consumers Union, the Center for Digital Democracy, and the Consumer Federation of America). At issue are considerations such as pricing, service defini- tions, interconnection terms, and rules for the use of public resources (e.g., radio spectrum and rights-of-way) as well as a variety of special needs that are usually met outside of normal market action (e.g., univer- sal service and public safety). Broadband is also associated with issues of privacy, security, and access by law enforcement; the complexity of these particular issues necessitates separate examination, and they are not dis- cussed in this report. Viewed through the lens of competition policy, the economic regula- tion associated with telecommunications policy is complemented by anti- trust and other matters associated with the structure and conduct of pro- vider industries. The action centers on administrative authorizations (with or without conditions) and legal decisions related to mergers and acquisi- tions. At issue are basic questions of market power and related conduct, such as interconnection and access to directories. Initiative; S. 966, Rural Broadband Enhancement Act. Bills that would support research include H.R. 2401, Rural America Digital Accessibility Act, and S. 430, Broadband Rural Research Investment Act. Bills that would change ILEC regulation include H.R. 1542, Internet Freedom and Broad- band Deployment Act; H.R. 1697, Broadband Competition and Incentives Act; H.R. 1698, American Broadband Competition Act; H.R. 2120, Broadband Antitrust Restoration and Reform Act; S. 1126, Broadband Deployment and Competition Enhancement Act; and S. 1127, Rural Broadband Deployment Act. S. 500, the Universal Service Support Act, would extend universal service fund coverage for broadband. S. 1056, the Community Telecom- munications Planning Act, would provide support for community planning grants. (Na- tional Journal’s Technology Daily. 2001. Broadband Bill Status. National Journal, Washing- ton, D.C. Available online at <http://nationaljournal.com/pubs/techdaily/briefroom/ billstatus/broadband.htm>.)

170 BROADBAND Viewed through the lens of consumer protection, legislative and ad- ministrative actions focus on consumer experiences. These include prices and quality of services, fair and clear billing, broadening of access to underserved areas and populations, linkage to other consumer concerns (e.g., education and health care services), and so on. And there are yet other perspectives. Because the greatest deploy- ment uncertainty and expected impact are in the last mile, broadband policy making at state and local levels looms large. It is at these levels that community and personal impacts are most evident. At state and local levels, economic development is increasingly linked to communications infrastructure because of expectations about how such infrastructure can contribute to economic opportunity (e.g., through skills and job develop- ment, attraction of employers, and support for telecommuting) and qual- ity of life (e.g., in terms of educational and health-care resources and community information sharing). Many states and a number of localities have taken a number of initiatives aimed at broadband deployment. Broadband is primarily an issue of domestic policy, but international per- spectives arise with respect to competitiveness issues and the interna- tional context in which the Internet exists. Recognizing that a diverse set of actors plays a role in shaping broad- band policy is important, because it is too easy to narrow the discussion to state and federal regulation of the price of communications services, a dominant historic concern in telecommunications. State and federal legis- lators, regulatory agencies, the courts, and local governments all influ- ence policy. Decision making takes many different forms: rule making, tariff setting, court cases, and voluntary agreements. The role of cable operators and companies using terrestrial wireless and satellite means that broadband policy encompasses more than traditional telephone rules. Relevant policy spheres include regulation of retail and wholesale rates; interconnection and unbundling rules; local cable franchising; access to poles, conduit, and other rights-of-way; wireless spectrum licensing; uni- versal service rules; and antitrust law. Finally, the committee notes that, much as is the case with other com- puting and telecommunications sectors, there are often valid conceptual arguments to be made on behalf of a number of different positions and interests in each policy debate. Which position represents the appropriate policy choice is a function of circumstances specific to the matter, but one rarely has the empirical evidence required to support one position defini- tively over others. Furthermore, the parties with interests in the outcome of a policy debate always represent their own positions as unambigu- ously right and true. Thus, at some level, one is ultimately working with educated guesses, and some mistakes are inevitable.

BROADBAND POLICY AND REGULATION 171 POLICY IMPLICATIONS OF TECHNOLOGICAL CHANGE The issues raised by the deployment of broadband technologies and services are in some ways quite different from the ones that policy makers face in regulating the narrowband telecommunications sector. In narrow- band, the principal issues facing policy makers today involve the fashion- ing of policies that will facilitate the efficient conversion of a mature mar- ketplace dominated by a single provider to a competitive marketplace served by different facilities-based and non-facilities-based providers. In the case of broadband, the marketplace is in its infancy and evolving in uncertain directions. Although major technological changes in the public switched telephone network have occurred, such as the movement from party-line to single-line service or the development and deployment of vertical services, the development and deployment of broadband services are occurring over a much shorter time. This leads to difficulties in esti- mated costs. Trends in the costs of some elements, such as customer pre- mises equipment, can be estimated on the basis of conventional wisdom about Moore’s law and the impact of rising production volumes. But every so often, some kind of component or process innovation can lead to a significant cost change.3 It is not surprising, given this young and evolv- ing marketplace, that the committee found uncertainty in predictions of per-subscriber costs and take-rates. At the same time, infrastructure tends to be enduring, which implies that errors in planning can have long-term consequences—a reality that has contributed to a relatively slow pace of regulatory evolution as well as conservatism in some segments of the telecommunications industry, even in the face of rapid change.4 Regulation in the Face of Rapid Change Historically, even when there was less-rapid technological change in telecommunications, governmental regulators generally have not at- tempted to control the process of change directly. The Federal Communi- cations Commission can exercise influence over technology for many pur- poses: to encourage or enable new services, which is the emphasis in this report; to assure certain protections (e.g., against radio frequency interfer- ence or safety hazards); and to support certain kinds of uses of the infra- 3For instance, companies have announced a number of innovations in the mechanics of deploying fiber, such as V-group splicing, blow-in of fiber into conduits, and robotic instal- lation in sewer pipes. 4For an earlier discussion of different industry cultures, see Computer Science and Tech- nology Board, National Research Council. 1996. The Unpredictable Certainty: Information In- frastructure Through 2000. National Academy Press, Washington, D.C.

172 BROADBAND structure that are given priority through public policy decisions (e.g., 911 emergency services, law enforcement access capabilities,5 and disability access). Early on, the government posture with respect to the market dominance of AT&T evolved to reflect evidence that AT&T devoted, with- out a government mandate but with protected revenue, considerable re- sources to improving the efficiency and capabilities of its network through technological advances. Later, government efforts to open previously closed markets motivated equipment manufacturers and service provid- ers to develop and deploy equipment and facilities that could take advan- tage of those commercial opportunities.6 Today, for example, literally scores of firms are deploying fiber-optic cables and advanced switches and routers in local and long distance networks to provide voice and data services, a phenomenon that (along with mobile wireless) has been fueled recently by tens of billions of dollars in venture capital. Rather than man- date particular technological solutions, the FCC has tended in recent years to address technology via selected performance requirements against which industry groups could develop specific standards, and even these activities seem to have diminished over time.7 In the turbulent wireless arena, the FCC has been changing its procedures for issuing radio li- censes, from adopting auctions for allocation of spectrum to considering a systemwide rather than site-specific licensing approach. Nonetheless, the FCC has been faulted for some of its approaches to new technologies.8 More generally, both regulators and investors have 5Communications Assistance for Law Enforcement Act, 47 USC 1001, PL 103-414. 6In the 1950s, for example, advances in microwave technology (originally developed for the government during World War II) created an alternative system for transmitting tele- phone calls over long distances, in lieu of AT&T’s embedded system of wires. In 1958, the FCC authorized large businesses to use microwave facilities to construct their own private networks. In 1969, the commission took the next step and permitted firms to compete di- rectly with AT&T for certain types of services. And in the 1980s, with divestiture bringing the realities of competition closer, AT&T executives came to revise their own assessment of the costs of adding fiber in their long-distance network as competitor actions, such as Sprint’s “pin drop” advertisements, made the case more compelling. 7In the case of “1+” access, for example, the FCC did not specify the particular types of modifications to existing telephone switching equipment that were required to provide “1+” access. Instead, it mandated the performance requirements that the carriers would have to satisfy and allowed the carriers, working with equipment manufacturers, to de- velop the specific technical modifications. 8Some argue that both the FCC and AT&T were slow to cultivate cellular telephony, where deployment and commercial service lagged key innovations considerably; that the granting of licenses for UHF television channels proved to be a costly diversion of resources and spectrum; and that the approach taken to standard-setting for advanced (“high-defini- tion”) television serves to slow progress in that arena. See, for example, “A Very Long Distance: A Regulatory Call Put Cell Phones on Hold,” Technology Review, May 2001, p. 110.

BROADBAND POLICY AND REGULATION 173 guessed wrong on many technologies; it has been easy for people to both over- and underestimate the potential of new technology. This point was emphasized by many participants at the committee’s June 2000 work- shop, which included comments on disappointments in such technolo- gies as Integrated Services Digital Network service and multipoint multi- channel distribution service (in its video distribution incarnation), among others, for which the technology proved limited in practice, penetration never reached anticipated levels, and better alternative technology was ultimately adopted. Such experiences underscore the fundamental diffi- culties that regulators have in gauging the coevolution of technologies and markets—and their influence on them.9 The problem of understand- ing the trends and implications of new technology seems especially acute for the Internet, since few regulatory agency staff, at least at the FCC, have been expert in Internet-related technologies.10 The FCC has recently launched several initiatives aimed at increasing its technical capacity in this and other areas.11 One of the most obvious indicators of the difficulty regulators (and other policy makers) have in keeping up with technology change is the definitions they develop and use. The Telecommunications Act of 1996 refers in general terms to “advanced services.” Required to report to Con- gress on deployment of advanced services, the FCC subsequently defined these to be at least 200 kbps in either direction.12 As discussed in Chapter 2 in this report, this sort of definition is problematical: 200 kbps will increasingly be, at best, a lowest common denominator in an environment 9At a June 2000 workshop, Thomas Krattenmaker of Mintz, Levin (and previously the FCC and academia) observed: “I would say that any regulation or any response you pro- pose to the FCC that is predicated on your ability to predict what technology will prevail, when, will be a useless recommendation . . . . We are just rife with suggestions, too many of which the Commission has adopted, that were based on some ability to know when tech- nology and which technology was going to be deployed. I don’t think we’re capable of knowing that, and I know the commissioners are not—they’re not selected on that [basis].” 10Casual observation shows that the FCC has engaged a single Internet-oriented indi- vidual in its Office of Plans and Policy since the mid-1990s, and beginning in the late 1990s it engaged chief technologists with Internet expertise, but there are limits to what a couple of specialists in staff positions can accomplish. 11Steps taken include the 1998 establishment of a Technical Advisory Council and the 2001 launch of an agencywide “Excellence in Engineering” initiative, including hiring and training measures. 12Broadband Second Notice of Inquiry, Federal Communications Commission (FCC, 2000, “Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996: Second Re- port,” CC Docket No. 98-146, FCC, Washington, D.C., August 21).

174 BROADBAND where capabilities of many technologies are growing; it is unclear whether bandwidth demand will be symmetrical, as the requirement assumes; and, in any event, the committee believes that use of any static definition is unwise over the long run. The separation, intellectually and culturally, of those setting policy definitions from those developing the technology may be unavoidable, but it has consequences. Thus, broadband is still a very new technology family, as is the Inter- net in general. The great and general potential of the technology seems clear, but how it might be effectively (and profitably) commercialized is still unclear. During this stage, one can expect to see much experimenta- tion with different business and service models by both private and pub- lic providers. It may be wrong to assume that executives whose public remarks exude confidence have deep understanding of all dimensions of the new approaches to networking that they are developing. For example, several people at the committee’s June 2000 workshop acknowledged in discussions that some of their earlier assumptions had proved wrong and that their views were evolving with experience.13 Indeed, the years 1999- 2000 saw the nation in the midst of a dot-com and telecommunications euphoria in the financial markets, while 2001 sees the reverse. All these observations, therefore, suggest caution in reading too much into the im- mediate situation and the importance of business strategists and policy makers staying as flexible as possible during this stage. Asymmetrical Regulation and Achieving Technology Neutrality The development of Internet-based services that can operate over both cable and telephone networks has accentuated convergence and tech- nology neutrality. Cable, DSL, and wireless providers can offer relatively comparable applications and services. Not only do these alternatives each offer high-speed access to the Internet, but each also has the potential to provide services that compete directly with the traditional offerings of other networks (e.g., cable broadband facilities can be used to provide voice services, and DSL can carry streaming video). The unsustainability of competition that may arise with asymmetrical regulation is most severe when two products or services are perfect sub- stitutes for each other. For example, if DSL and cable modem services are 13For example, in remarks to the committee, Sprint’s Jim Hannan said of his MMDS offering, “We don’t have effective models . . . so we really don’t understand how the net- work behaves. We’re pushing it every day.” Hannan observed that projected upstream-to- downstream traffic ratios were much higher than what was observed when Sprint de- ployed its network; this was attributed largely to customer use of Napster.

BROADBAND POLICY AND REGULATION 175 extremely close substitutes, disparate rule sets are likely to be a signifi- cant problem in the long run, unless bundling broadband access with other services (cable television or telephone) creates some differentiation. To the extent they are differentiated in ways that reflect demand hetero- geneity, this is less of a problem. Still, even if the platforms do not provide directly compatible services, they may provide services that are nonethe- less substitutable for each other. Their convergence notwithstanding, these technologies are for his- torical reasons subject to separate and substantively different regulatory regimes—a situation characterized by some academics, industry repre- sentatives, and FCC officials as regulation in “stovepipes.” At the FCC, stovepipes are embodied in bureaus—cable services, common carrier, mass media, and wireless telecommunications. Those bureaus, in turn, correspond to focused enabling statutes, which reinforce the traditional linkage of technologies with industries and specific services. Appendix B contrasts the regulation of telecommunications common carriers (which affects DSL provision) with that of cable operators. Terrestrial wireless and satellite are subject to yet another set of distinct regulations. Looking across the various stovepipes, the greatest constraints appear to apply to the telephone industry, where rules (retail price regulation and newer market-opening requirements) are intended to inhibit ILECs from taking unfair advantage of their historically dominant position as regulated mo- nopolies in telephony as they enter into other market segments; but one also sees cable franchising provisions being applied to broadband deliv- ered over cable. The issue of asymmetrical regulation of broadband has been high- lighted by the current debate over “open access,” discussed in detail be- low. Cable operators have maintained that the Internet access service offered over their networks is a cable service and, consequently, that they are not required to offer unaffiliated ISPs that wish to reach cable sub- scribers access to this service. Opponents have claimed that cable opera- tors are engaged in the provision of a telecommunications service when they offer high-speed access to an ISP and, hence, sought to have regula- tors require cable operators to offer that service on a nondiscriminatory basis to unaffiliated ISPs. (This argument has had traction in the courts; one response has been a move by some cable companies, which have resisted being classified as telecommunications services, to embrace the designation as a way of avoiding franchise payments, illustrating the maneuvering that goes on.) Other parties have contended that govern- ment intervention is unnecessary (and may be harmful), because market- place forces are sufficient to cause cable operators to make access avail- able to unaffiliated ISPs as soon as technical problems are addressed and business arrangements supporting access to multiple ISPs are put in place.

176 BROADBAND While technical development continues, successful demonstrations and availability of unaffiliated ISPs in some markets suggest that the technical problems of supporting multiple ISPs in cable systems can be overcome.14 Business aspects—such as provisioning and troubleshooting systems— are a focus of current activity. Further confounding the question are technology trends suggesting that fiber will be pushed ever deeper into wireline and terrestrial wireless broadband networks, meaning that provider networks will increasingly resemble each other even in terms of technology, at least in all but the last segment of the access link. As the similarity in both product (services) and delivery technology grows, the distinction between the service-produc- ing and service-delivering industries falls, but whether or how one recon- ciles the associated regulatory regimes is an open question. Regulatory legacy and convergence aside, some regulatory issues speak to inherent attributes of a technology, confounding any notion of a technology-neutral policy. Implementing local loop unbundling, for ex- ample, involves an intimate understanding of the physical environment of that plant. Policy with respect to public rights-of-way involves the details of poles and conduit access for wireline systems and tower siting for wireless. Wireless illustrates another challenge to achieving technol- ogy neutrality: with different services—including fixed terrestrial service such as MMDS and mobile services—assigned to different frequency bands, allocating scarce spectrum among these would seem to require making technology- and service-specific trade-offs. In some cases, it is entirely reasonable to make distinctions: one does not have spectrum auctions on fiber, for example. However, when obligations of any kind are imposed—performance obligations certainly—it is generally very dif- ficult or impossible to be completely technologically neutral. How can one treat these services on an even-handed basis despite the differences in the underlying technology? The committee’s assessment of technology options (see Chapter 4) suggests that broadband will not involve a technological horse race or overall regulatory or market choice among technological options. The various access technologies will fill different niches, and multiple connec- tions will be available in many markets (many already have two connec- tions—phone and cable lines—capable of supporting broadband to the 14For example, in June 2001, AT&T announced that its Boulder, Colorado, open access trials were successful (Richard Williamson, “AT&T Completes First Open Access Cable Trial,” Interactive Week. June 7). Earthlink subscriber information as of October 2001 indi- cates that its services are available over Time Warner Cable systems in several markets and that more markets will be added in the near future.

BROADBAND POLICY AND REGULATION 177 home). This diversity suggests that the issue of technology neutrality will be around for some time to come. To the extent that neutrality is not achieved, regulatory actions would favor or disfavor options in ways that could decrease investment incentives or otherwise distort natural market forces in ways unfavorable to consumers. Decreased choice would reduce the likelihood that facilities-based competition emerges or would deprive consumers of particular cost and performance options. The existence of these asymmetries, as well as the looming contradic- tions that convergence in the access technologies themselves poses, is recognized by regulators. What is less clear is how to craft a solution that improves the situation. One problem may lie in the concept of neutrality and regulators’ goals not to pick winners and losers: neutrality may not be feasible, given the tight coupling between the subjects of regulation and the details of particular technologies. In avoiding tying regulations to technology-specific considerations, one has to find some other space— some more abstract definition of broadband—in which to regulate. Here one is confounded by such factors as the dynamic nature of broadband and the two-way link between technology-specific performance charac- teristics such as speed and application requirements. Also, absent the specific issues that have been the source of much of telecommunications regulation, what would be the goals of regulation of “broadband”? COMPETITION The establishment of robust competition among multiple telecommu- nications providers, including broadband and other providers, is a basic premise of the Telecommunications Act of 1996 (Box 5.1). This is viewed by many as the desirable way of making broadband as affordable as possible, though the view is not universal.15 Two principal paths toward competition are contemplated in the present policy regime—(1) unbun- dling and resale and (2) facilities-based competition—which are discussed in separate subsections below. Unbundling arose in the context of policies aimed at stimulating competition to the ILECs. The 1996 act mandated unbundling of local loops and other network elements. In contrast to unbundling, facilities-based competition involves new entrants using their own equipment and physical network to compete. 15For example, the National Telephone Cooperative Association (NTCA) commissioned a white paper that concluded that the entry of competitors would decrease the take-rate achievable by any single carrier, which could substantially undermine the financial case for DSL in rural areas where it is already constrained even without competition. NTCA linked the issue to its call for only incumbents to be eligible to receive universal service high-cost- area support payments (Telecommunications Reports, January 15, 2001, p. 6).

178 BROADBAND BOX 5.1 Key Provisions of the Telecommunications Act of 19961 Related to Broadband • Section 251 of the Telecommunications Act of 1996 establishes a series of obliga- tions that apply to telecommunications carriers. Some apply to all telecommunications carriers (local as well as long distance and others). Some apply only to providers of local telephone. The most detailed requirements apply to incumbent local telephone compa- nies, such as the Bell Operating Companies. The latter—those pertaining to incumbents— consist of a variety of obligations that collectively are designed to facilitate the entry of new providers into local markets and enhance their ability to compete with the incum- bents. These include, for example, a requirement that incumbents make available parts of their local networks to competing providers on just, reasonable, and nondiscriminatory terms and conditions. The procedures for implementing these requirements for incumbent telephone companies are set forth in Section 252. • Section 253 generally preempts, with certain limited exceptions relating to univer- sal service and other public policy objectives, any state or local statute or regulation that prohibits or has the effect of prohibiting the ability of any entity to provide any interstate or intrastate telecommunications service. • Section 254 promotes access to advanced telecommunications and information services in all regions of the nation. Universal service principles to be implemented by the Federal Communications Commission include ensuring the following: quality services at reasonable and affordable rates; access to advanced services; access to such services in rural and high-cost areas; that all providers of telecommunications services make an equi- table and nondiscriminatory contribution to the preservation and advancement of univer- sal service; that specific and predictable support mechanisms are in place to carry out such preservation and advancement; that there is access to advanced telecommunications services for schools, health care, and libraries; and that other principles that the joint (federal-state) board and the FCC may determine are necessary and appropriate for the protection of the public interest are implemented. • Section 255 requires telecommunications products and services to be accessible to people with disabilities. This is required to the extent that access is “readily achievable,” meaning easily accomplishable, without much difficulty or expense. If manufacturers can- not make their products accessible, then they must design products to be compatible with adaptive equipment used by people with disabilities, where readily achievable. What is “readily achievable” will be different for each manufacturer, depending on the costs of making products accessible or compatible and their resources. • Section 256 sets broad parameters to establish nondiscriminatory access for the broadest number of users and vendors of communications products and services to the public telecommunications networks that are used to provide telecommunications service through joint network planning. It defines “public telecommunications network intercon- nectivity” as the ability of two or more public communications networks used to provide telecommunications service to communicate and exchange information without degener- ation, and to interact in concert with one another. This section also regulates coordination for interconnectivity and establishes FCC procedures for oversight. It sets out the parame- 1Formally, the Communications Act of 1934, as amended.

BROADBAND POLICY AND REGULATION 179 ters under which the FCC is to review and eliminate federal regulations that may act as market-entry barriers for entrepreneurs in providing telecommunications. The FCC was required to conduct such an initial proceeding within 15 months of the law’s enactment and thereafter to conduct similar periodic reviews every 3 years. • Section 259 mandates that incumbent local exchange carriers (ILECs) make avail- able to any qualifying carrier any public switched telecommunications equipment or in- formation as should be requested by the qualifying carrier. It excepts situations under which it would be economically unreasonable or against the public interest for the ILEC to comply. It permits joint ownership and seeks to ensure that the ILEC is not treated as a “common carrier for hire” and that the carrier seeking the use of facilities will be allowed the use of these facilities on just and reasonable terms. Finally, Section 259 demands a transparent process, requiring the ILEC to report the terms and conditions of any facilities- sharing arrangements. • Section 271 requires that the FCC consult with the U.S. Department of Justice and the relevant state commissions before ruling on a Bell company’s request to offer in-region interLATA services. Upon application by a Bell company, the FCC has 90 days to consider whether the applicant has met a 14-point “competitive checklist” of market-opening re- quirements contained in the section and whether the company’s entry into the interLATA service market is in the public interest. • Section 301 stipulates that the FCC shall review any complaint submitted by a franchising authority concerning an increase in rates for cable programming services and issue a final order within 90 days after it receives such a complaint, unless the parties agree to extend the period for such review. • Section 302 eliminates the prohibition on local exchange carrier (LEC) provision of video programming in the LECs service area. LECs and others may offer video program- ming under regulations that vary according to the type of video service being provided (radio-based, common carriage, cable TV systems, or open video systems). With the law’s enactment, regulation was lifted for cable programming for a basic service tier that was the only service subject to regulation on December 31, 1994, in any franchise area in which the operator serves 50,000 or fewer subscribers. Acquisitions and joint ventures (Section 302 of the bill, Section 652 of the act) are to a large extent prohibited, though there are several exceptions for certain small and rural systems. The law also permits a LEC to acquire or joint venture under different terms and condition in cases where the subject market meets the FCC’s definition of “competitive.” The FCC may waive the acquisition and joint venture prohibitions if it determines that the economic viability of the market merits such or that to do so would otherwise be in the public interest, and if the local franchising authority approves. • Section 303 allows cable operators to provide telecommunications services with- out first obtaining a franchise to provide those services. Additionally, no franchising au- thority may interrupt a cable operator’s telecommunications services based on that oper- ator’s lack of a franchise. The section also prohibits franchising authorities from requiring that any cable operator provide telecommunications services as a condition for granting a franchise. • Section 706 seeks to promote the deployment of “advanced telecommunications services” in a reasonable and timely fashion. It attempts to do this by means of price continues

180 BROADBAND BOX 5.1 (continued) cap regulation, regulatory forbearance, measures that promote competition in the local telecommunications market, and other regulating methods that remove barriers to infrastructure investment. This section also required the FCC to follow up with inquiries into the progress of deployment. Reports issued in August of 1999 and 20002 found deployment reasonable and timely based on subscribership levels, ser- vice and technology options, and infrastructure investment at the time of the inquir- ies. The August 2000 report observed that advanced services may be unevenly dis- tributed owing to differences throughout the country in wealth and population concentration. 2Federal Communications Commission (FCC), 1999, “Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fash- ion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecom- munications Act of 1996: First Report,” CC Docket No. 98-146, FCC, Washington, D.C., August; and FCC, 2000, “Inquiry Concerning the Deployment of Advanced Telecommunications Capa- bility to All Americans in a Reasonable and Timely Fashion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996: Second Report,” CC Docket No. 98-146, FCC, Washington, D.C., August 21. Unbundling and Resale Mandates Unbundling refers to the breaking down of an incumbent’s network into smaller subcomponents, which can be either technology components (e.g., a phone line) or service components (e.g., switching), so that these elements can then be sold separately to other service providers. The goal is to permit new competitors to compete with the incumbent without having to incur the costs and the risks of constructing all of these elements themselves. An important difference between resale of services and physi- cal unbundling of network elements is how much leeway the competitor has for differentiation. With simple resale, the competitor is confined to deriving revenue from the differential between the resale and retail rates, whereas unbundling gives the competitor latitude to provide differenti- ated services that combine unbundled elements with elements provided by the competitor. Most prominent in the context of broadband deployment is unbun- dling of the local loop. The ILEC local access facilities have been the subject of unbundling rules designed to enable CLECs to offer voice and data services without having to build their own local access facilities. The DSL competitor provides the facilities at both ends of the loop (the DSL modem and DSLAM) and connects them to the actual copper wires that

BROADBAND POLICY AND REGULATION 181 make up the loop. Under present rules, the CLEC has considerable free- dom to select the particular DSL technologies (and thus such parameters as speed, ratio between up- and downstream speeds, and the maximum loop length supported), including technologies not offered by the incum- bent. CLECs have argued that such differentiation is critical, because the incumbents have generally not deployed the high-rate, symmetric DSL services favored by small businesses (one argument is that this reflects incumbent reluctance to cut into the more costly, profitable T1 data ser- vices offered by the incumbents). This form of unbundling, in which a passive network element—the copper loop—is made available in raw form to the competitor has the additional advantage to the competitor of helping to isolate the quality and nature of the competitor’s service from potential adverse actions of the incumbent (though issues of successfully provisioning the loop in the first place and dealing with crosstalk within the cable bundles remain). As an alternative to physical unbundling of copper loops, the ILEC could also be required to provide access to its DSL at the packet level (through ATM or IP technologies). Cable has a different context for unbundling, both legally and techni- cally. Owing to the design of today’s cable systems, most notably the shared communications medium, strict unbundling along the lines of loop unbundling for DSL is not practical. The open access arrangements con- templated by the FCC’s order in the AOL Time Warner merger (and similar arrangements being explored by other cable operators) lie some- where closer to resale than to physical-layer unbundling. The cable opera- tor operates the cable system over which the unaffiliated Internet provid- ers connect to customers and hands off the packets destined for the unaffiliated providers at the cable system head end. Under the terms of the AOL Time Warner merger order, the cable operator must not dis- criminate against the unaffiliated provider on technical quality; by the same token, the unaffiliated provider does not have latitude to compete with AOL Time Warner on the basis of the technical quality of the last mile connection. Resale and unbundling rules raise a variety of concerns on the part of the facilities owner and the competitor. When a competitor is dependent on another for critical inputs, the facilities owner will have the incentive and perhaps the ability to use its control over the input to disadvantage the competitor in the downstream market in which the firms compete. Further, as discussed above, facilities owners express the concern that they will never be fully compensated for their costs under regulator-set access prices, and how such prices should be set is a matter of current debate. Not surprisingly, these issues have been vigorously debated since resale of telephony began, and they account for a significant fraction of telecommunications regulatory proceedings and related court cases.

182 BROADBAND When Unbundling Works Unbundling has been playing an important role in broadband com- petition. CLECs have provided service in areas unserved by the incum- bents, and some credit the CLECs for having stimulated deployment ef- forts by the ILECs. As discussed above, competitive service providers have entered the wholesale DSL business by leasing local loops and colocation space from ILECs. However, CLECs’ long-term impact on the competitive landscape is in doubt. The competitive DSL industry faces an uncertain future at present, reflected in a series of major business failures. These problems reflect the difficulties inherent in forcing an incumbent monopolist to open its market to competitive entry, the effects of econo- mies of scale and scope on smaller players, and, from the vantage point of 2001, the challenges in raising investment capital for any sort of tele- communications investment. These uncertainties raise questions about how to think about unbundling in fashioning future policy. Unbundling creates an enforced market structure out of the assets of an incumbent. One should ask how efficient the resulting market could be—how efficient and effective are control and management with mar- kets based on unbundled versus integrated industry structures, and what factors facilitate unbundling? The basic thrust of the literature on organi- zational factors is that, when technologies are easy to understand and simple, markets are likely to have a comparative advantage as coordina- tion mechanisms owing to the incentive effects of competition.16 In this circumstance, unbundling can be effective. But if the interfaces between players are complex, an integrated business approach will often be more effective. The technical complexity associated with unbundling for broadband access has been seen in the case of both DSL and cable. In the case of cable video programming, there is a high-level unbundling today in the form of certain channels’ being set aside for use by local broadcasters; public, educational, and government access; and leased access. These are logical unbundling variants with simple technical implications. But more recent proposals have focused on an unbundling at the Internet Protocol layer. (There were also early proposals to provide unbundling through the use of asynchronous transfer mode, or ATM, technology.) Concerns today center on the optimal technical mechanisms to support shared access and on establishing operational and management mechanisms for coordina- tion (provisioning, troubleshooting, and the like) between the cable com- pany and heretofore unaffiliated ISPs. 16See, for example, Oliver E. Williamson, 1975, Markets and Hierarchies, The Free Press, New York.

BROADBAND POLICY AND REGULATION 183 As discussed above, DSL unbundling can occur at the physical level, which permits variation in the DSL service that the competitor provides, and at the resale level. Where ILECs extend fiber closer to customers, replacing a portion of the existing copper plant, copper pairs no longer run all the way from premises to the central office, making physical un- bundling very complex. Issues that must be solved include whether and how CLECs should be given access to colocation space in remotely de- ployed pedestals, equipment vaults, or even equipment located on pole tops. The incumbents have claimed that this level of complexity inhibits investment in new facilities and is a barrier to the progress of broadband deployment. In turn, CLECs are concerned that installation of new facili- ties such as fiber-fed remote terminals would complicate or preclude the physical loop unbundling on which their businesses depend. In addition to making colocation and unbundling more complex, remote terminals alter the economics of physical-layer unbundling, because competitors are able to serve far fewer customers from a single colocation point. A specific example of this debate centers on Project Pronto, in which local exchange carrier SBC Communications has proposed a program of fiber deployment that would enable it to provide high-speed digital ser- vices to its customers that are being served by long copper loops or with loops fed from digital carrier loops, and that as a result cannot currently obtain DSL service. Because the program as originally described would have deprived the CLECs of the opportunity to select the equipment at each end of the remaining loop segments, they argued that the program would make it difficult for them to use the new infrastructure to do any- thing other than resell SBC’s DSL service, meaning that they would be unable to differentiate their own service in order to compete with SBC. SBC argued that unbundling at a remote terminal in DSL at the physical layer is so costly and complex as to be impractical, making higher-level unbundling solutions a better alternative. The FCC preferred physical- layer unbundling, making the plan’s approval contingent on SBC’s agree- ment to increase the size of the remote terminals to permit competitors to install termination equipment there rather than simply to resell SBC’s service. The issue of incumbents’ unbundling obligations when they de- ploy remote terminals is currently the subject of a pending FCC rule- making proceeding. In the case of DSL, another major complication is that the individual wires connecting subscribers to central offices cannot be treated as en- tirely independent of each other, as physical unbundling would imply. As detailed in Chapter 4, crosstalk, the coupling of electrical signals be- tween nearby wires, creates interference that can degrade the carrying capacity of each copper pair, decreasing the data rates supported by a given line length and decreasing the maximum line length over which

184 BROADBAND DSL can run. The interference effect depends on how the signals on the different pairs make use of the different frequencies used for transmission over the lines, which in turn depends on how the ILECs and CLECs make and coordinate decisions about provisioning customers whose twisted pairs share bundles. The problems will grow as the penetration of DSL grows and as the higher data rates contemplated in the DSL upgrade path—which are still more sensitive to interference—begin to be widely implemented. Coordination may be facilitated by technical standards, which are in development, but which would, if adopted on a mandatory basis, constitute a regulation of operational network technology. Such a standard aims to achieve the sort of clean technological interface alluded to above, thus reducing coordination problems. However, the more the coordination problems are solved, the less the flexibility of the competitor to differentiate its service, so in the end, physical-layer unbundling may be no more attractive then simple resale, though much more complex. Implications for Investment by Incumbents There are major incentive issues that bear on expectations for overall investment and innovation. The analytical models used to justify forcing incumbents to unbundle and sell access to network elements implicitly assume that incumbents’ networks are based on a static technology and involve only facilities already deployed. In reality, networks are con- stantly being upgraded, and with the upgrades come new capabilities and services. Competitors argue that incumbents should unbundle new services and technologies and resell them, as they are required to do with old services and technologies. Incumbents argue that they have no incen- tive to invest in new facilities or otherwise innovate if they are forced to sell their innovations at cost to their competitors. In particular, where unbundling is mandated at regulated prices, the incumbent bears the risk of investment but cannot fully benefit from it. Facilities-Based Competition Under facilities-based competition, competitors go head-to-head, us- ing independently built and operated local access infrastructure.17 It is widely believed by economists, policy officials, and consumer advocates 17Facilities-based competitors may still make use of some facilities such as backhaul cir- cuits that are owned by other telecommunications companies, including the ILECs, and all facilities-based competitors must at some point interconnect with the other ISPs that make up the Internet.

BROADBAND POLICY AND REGULATION 185 that facilities-based competition is the preferred end state (with the pro- viso that it also results in a reasonably efficient result, which may depend on the characteristics of the technology or market). A principal argument is that facilities-based competition is the only circumstance that will per- mit complete deregulation of local markets. In this view, competition though local loop unbundling or resale is a transitional approach to be used while facilities-based competition is still developing. As new entrants grow and gain market share, they will find it economic to replace facilities leased from incumbents with facilities leased from nonincumbents or self-provided. (The cost of using an incumbent’s facilities is not just the price assessed for the network elements used, but the ongoing difficulties associated with relying on a dominant competi- tor.) In essence, resale and unbundling rules do not remove the need for regulation, only shifting it from regulation of end customer prices to regu- lation of prices charged by facilities owners to reseller-competitors. As long as competitors are dependent on incumbents for some facilities, such as loops, regulation of the terms and conditions of the competitors’ access to those facilities, including price, will be required. Competitors are truly independent of each other only if they have their own facilities (or access to comparable facilities that are not controlled by a dominant competitor). There is a fundamental tension between the short-run static efficiencies of unbundling or resale and the longer-run dynamic efficiencies of facilities- based competition. Forced unbundling or resale at regulator-mandated prices may permit competitors to deploy innovative new services. How- ever, such measures also could lock in the current situation, undercutting the longer-term goal of full facilities-based competition, especially if the rule is that competitors will be granted access at controlled prices to any new facilities that an incumbent puts in place. Structural Separation Unbundling and resale mandates are among a range of interventions that could be invoked to address the market power of incumbent tele- phone companies by facilitating competitors’ access to upstream inputs controlled by the incumbent. Several options would involve some sort of separation of the incumbent’s lines of business. The weakest form, non- structural separation, involves the use of accounting safeguards to sepa- rate wholesale and retail operations. The most extreme mechanism, di- vestiture, involves the spin-off of a unit into a separate corporate entity that is under different control. Structural separation, between these two in terms of the level of intervention, involves the separation of business units into distinct corporate entities that remain under common control. That is, an ILEC would be required to offer access to its facilities and

186 BROADBAND services at nondiscriminatory wholesale rates to nonaffiliated retail ser- vice providers as well as to its own affiliated providers.18 Structural separation does not necessarily reduce the incumbents’ in- centives to discriminate against competitors or cross-subsidize, because the two affiliated entities remain under common control. That is, the firm as a whole still benefits if the wholesale entity discriminates in favor of its retail affiliates and against unaffiliated retail competitors. (The incentive to discriminate and cross-subsidize is only fully eliminated if retail opera- tions are divested from wholesale.) But structural separation may dis- courage the incumbent from engaging in such behavior by making the transactions between the wholesale and retail entities easier for regulators and other outsiders to monitor. To the extent that the structural separation approach succeeds, it has the potential drawback in the long term of decreasing the incumbent’s incentive to upgrade the telephone plant because its ability to fully cap- ture the benefits of that investment is constrained. Another potential drawback of structural separation is that, if there are valid reasons for vertical integration (e.g., reduction of transaction and/or coordination costs), then separation could result in net higher costs for consumers. Past history with unbundling and interconnection rules also suggests that sig- nificant regulatory effort could be required to carry out structural separa- tion. Finally, application of separation rules to the ILECs would also com- plicate efforts to reconcile policy across broadband delivered by different technologies or sectors of the telecommunications industry. How Much Competition Is Enough? Will facilities-based competition prove sufficiently robust to permit increased deregulation of broadband and to forestall the need for future retail regulation? If robust facilities competition were achieved, it might be possible to back away from existing unbundling mandates,19 and it might remove the impetus for initiating new mandates in the future (such as those contemplated under the rubric of cable open access). In the case of broadband, technology options that provide a founda- tion for facilities-based competition are at hand. Two wireline technolo- 18This is somewhat analogous to the separation that has occurred with electricity deregu- lation in a number of states, whereby the electric utility’s generation and distribution opera- tions are separated and customers can choose which generation company to purchase power from. 19The FCC has an unbundled network elements (UNE) remand process for removing network elements from the list of required elements to be unbundled if they can reasonably be self-provisioned or are available in the market.

BROADBAND POLICY AND REGULATION 187 gies—DSL and cable modems—that make use of distinct preexisting com- munications networks—telephone wires and hybrid fiber coax cable sys- tems—are being deployed on a wide scale. The networks that they build on provide access to the vast majority of the U.S. residential population, though the cost and feasibility of upgrading to support data services vary from location to location. A limited amount of facilities-based competi- tion from new entrants is also visible. Companies are investing in “over- build” of incumbent cable providers to compete with them, primarily using HFC. Two wireless technologies—terrestrial service using micro- wave transmission and two-way satellite service—are being developed and deployed as well. The long-term economic viability of these competi- tors is uncertain at present. The bottom-line issue with respect to evaluating competition is mar- ket power—whether a company can keep its product price significantly above the competitive level. Associated with this are concerns that a com- pany facing insufficient competition will let reliability and quality suffer or will fail to deploy new services. Market power is assessed in terms of characteristics of the market—how many competitors there are, what their relative market share is, what the ability of competitors is to expand out- put rapidly (using either current capacity or capacity that can easily be expanded), and how easily new firms might enter the market—and indi- vidual firm conduct (whether it is anticompetitive). Some guidelines do exist, such as those developed by the U.S. Department of Justice and the Federal Trade Commission (FTC) for use in evaluating proposed merg- ers, but the analysis in a merger situation is potentially different from that involved in regulating an ongoing market. More generally, there is no firm basis in economic theory to say how many competitors are “enough” in the abstract or across all markets or in all circumstances, and opinions vary considerably. For example, in remarks before this committee, repre- sentatives of consumer groups argued in the context of cable open access that a choice among four providers might well be insufficient. Evolution in technology or business strategy could significantly change the nature and terms of competition. For example, even if broad- band access is itself reasonably competitive, a shift toward exclusive bun- dling of content and search services with broadband access could lead to the disappearance of stand-alone ISPs as broadband becomes the domi- nant mode of Internet access. The supposition today is that the various broadband technologies provide a similar service, an Internet-based plat- form. However, both the technologies and associated business models are malleable, so it will be important to track their evolution and interactions over time to understand whether the different access technologies con- tinue to be reasonable substitutes for each other, or whether technological development or business strategies have caused them to diverge. Such

188 BROADBAND forces have the potential to lead to a more tiered market—or distinct markets—in which different technologies and industries are associated with different kinds of service (e.g., different bandwidths, applications, or prices). Finally, consumers with only one or two providers may derive spill- over benefits from competition in markets with more competitors, be- cause marketing and pricing programs often do not have sufficient granu- larity to discriminate in terms of price or quality between market type and because of negative public reactions to highly differential pricing. Assessing the Degree of Competition Precise data are limited, but the deployment numbers presented in Chapter 1 of this report suggest that facilities-based competition in broad- band is beginning to occur in the United States, with ILECs and cable operators undertaking large-scale deployments in many locations across the nation, and overbuilders entering a handful of markets. Wireless is an alternative in several test markets, and satellite services offer another option. However, overall availability masks considerable variability in competition at a local level—by state, by community, or even by house- hold. It is a yet-unanswered but critical empirical question whether broad- band local access will turn out to be a natural monopoly (as telephony was assumed to be for many years) in some or all markets. If so, it may continue to be dominated by the incumbent telephone companies and cable system operators, limiting facilities-based competition to at best two players. The fact that facilities-based competition has proved difficult to establish in the voice telephony markets that were the primary focus of the 1996 act, especially for residential service, is not encouraging, but there are differences between entry into a mature, saturated market and a new, evolving one. At this point in time it is hard to conclude what the overall shape of the market will be. It is, however, apparent that there are some geographi- cally defined markets that give rise to concerns about whether broadband service will be available within several years and whether there will be an adequate level of competition. In hard-to-serve areas, the sheer costs or business risks may be great enough to call into question the goal of creat- ing a competitive market. In these areas, the key goal may be providing service at all. In deep rural areas, where costs are highest, it remains to be seen whether the performance capabilities, costs, and ability to scale up of the existing or planned satellite services will be sufficient to keep custom- ers in deep rural areas at rough parity with others with regard to broad- band access. Both the extent of coverage (in terms of geographical regions

BROADBAND POLICY AND REGULATION 189 and households) and the level of competition will be the subjects of con- tinuing scrutiny by public policy makers and other interested parties. This suggests the ongoing importance of solid data, collected on a system- atic basis, to identify where and what level of competition is being cre- ated. Open Access and Evolving Complements to Facilities-Based Competition When cable operators began upgrading their systems to provide cable-based broadband services, concerns were raised that they could leverage their established cable television infrastructure and franchises to exercise market power in broadband services. The position of incumbent telephone companies could inspire similar concerns, but they have been required to provide would-be competitors with access to their facilities (via unbundling or resale), and public attention has been focused on the cable system operators. The open access debate, which began in 1999, has catalyzed involvement of consumer advocates, who had had a lower pro- file in early- and mid-1990s debates about Internet policy, and has pro- vided an important complement to traditional regulation in shaping broadband policy. Critical actions were the merger oversight by the FTC as well as the FCC and court decisions reviewing local efforts to require open access. One aspect of the open access debate has been implementation—how to implement open access, what the actual extent of proposed technical difficulties is, and how much the additional costs of supporting competi- tive ISP access over cable facilities would be. When cable modem systems were first developed in the mid-1990s, little thought was given to provid- ing outsiders access to the systems or technology to implement such ac- cess. The technical specifications were developed through industry stan- dard setting, coordinated by Cable Laboratories and building on company contributions of know-how. Cable Labs has since evolved those stan- dards, which have become more supportive of open access, but its cable industry orientation—which contrasts with the more open, broad Internet Engineering Task Force that developed many of the Internet’s core stan- dards—has caused it to be viewed with suspicion by competitors and consumer advocates. In light of the growing attention to open access, cable operators have been exploring technical options for supporting ac- cess by multiple ISPs. The goal of opening up an incumbent’s facilities could, in principle, be achieved in several ways, either at a low level by unbundling the physical links of the provider, or by unbundling some higher-level service. Early on, low-level access to cable infrastructure— for example, allocation of different frequency bands to different provid-

190 BROADBAND ers—fell out of favor, and logical-layer access schemes, in which packets would be routed to and from the appropriate ISP at the cable system head end, have been adopted. There is general agreement that open access can be implemented, and several pilot efforts have been conducted. A central feature in the open access policy debate has been uncer- tainty about business models for broadband—how will cable providers behave, and what do consumers want? In adding Internet services, cable operators have had to evolve both their infrastructure and business mod- els. At least initially, cable operators have tended to select a single ISP to provide Internet service over their facilities, and the service has typically been bundled with at least some content or services aimed specifically at their own customers. This reflects in part an extension of usual cable industry business practices to the new technology, but also the greater ease that having a single partner can provide when entering a new, evolv- ing market. In contrast, as critics frequently observe, much of the dial-up ISP market has emphasized unrestricted access to Internet content (AOL, which grew out of a closed content model, is the most notable exception, but several others, such as MSN, have also based their strategies on a mix of proprietary and Internet content). The open access debates have been informed by attitudes (positive and negative) about entertainment industries in general (which are asso- ciated with content generation and delivery) and television in particular. Concerns have been expressed that cable operators will seek control over the Internet content available to their subscribers in a fashion analogous to what they exercise with conventional television content. One possibility is that particular content would be favored, to the disadvantage of content from unaffiliated major content companies, niche sources, or nonprofit organizations, through placement on start-up screens, differentiated quality of service, or other means. Another is that cable operators would go one step farther and create a “walled garden” service in which outside content is difficult or impossible to access.20 The 20A Cisco white paper (Cisco Systems, Inc. 1999. “Controlling Your Network—A Must for Cable Operators,” available online at <http://www.cme.org/access/broadband/ cisco_MSOs_white_paper.html>) promoted technology for cable companies to control, perhaps preferentially, content flows from different providers. The white paper addresses cable operators’ concern about bandwidth-hogging traffic from other content providers and service quality in a context where the operator has arranged for supply of certain known content and customer use of other content is unknown but could be large, while in the short term network capacity is finite. It speaks to cable operators’ concerns about resource management by noting that Internet technology permits network resources to be managed. “Sustained service quality over the long term requires IP network control, be- ing able to intelligently segment and manage resources by user type, service, destination, or application so that delivery quality does not suffer with growth or the addition of new services” (p. 2).

BROADBAND POLICY AND REGULATION 191 term “walled garden” arose in the ostensibly benign if parochial context of cable marketing and of cable literature from the late 1990s. It has been cited by consumer advocates as evidence of adverse intent,21 though the kinds of preferences that may arise from a walled garden strategy can also be viewed as similar to what has been seen elsewhere in television and newspapers, namely, the use of a transaction with a partner to gener- ate revenue that can lower the price charged to consumers and the pre- selection of a menu of content deemed to be of greatest interest to con- sumers. A walled garden is not, per se, hostile to consumers, because it may lead to lower prices and increased content. The concerns about re- stricted choice or walled gardens stem in part from the contrast with the unfettered, unfiltered access to content throughout the network that has traditionally characterized Internet services. Additionally, there are con- cerns about media concentration and the relative market position of broadband providers that also operate content businesses. Ultimately, whether and how consumers can access content from other sources will depend on the industry response to consumer demand (and the condi- tions attached to industry acquisitions and mergers) unless public interest is deemed important enough to mandate such access. There are also concerns that those cable operators who are in a mo- nopoly position in their market, which is the case in most markets, will be able to charge other ISPs or content providers access fees that are higher than the costs charged to ISPs or content providers affiliated with the cable operator, giving the latter a competitive advantage. Concerns are greatest for those affiliations that involve ownership (e.g., AT&T’s inter- est in @Home or the AOL Time Warner combination). The ILECs, which are required to unbundle their lines to permit resale competition, have expressed the view that not imposing such requirements on cable repre- sents an unfair advantage arising from asymmetrical regulation. Cable companies respond that open access requirements could deter them from investing in system upgrades to support broadband if they perceive the likely returns under open access conditions—conditions simi- lar to those faced by ILECs—to be insufficient. These business judgments are commonly invoked as threats in a political process; they are, of course, impossible to verify ex ante. In 2000-2001, the nation’s two largest cable system operators agreed to provide customers a limited choice of ISPs. The rules under which this will be done were outlined through conditions placed on the AOL Time Warner merger by agreement with the FTC in 2001, which will be moni- 21See, for example, Center for Media Education, 2001, Broadband Networks and Narrow Visions: The Internet at Risk, CME, Washington, D.C., available online at <http://www.cme. org/access/broadband/at_risk.html>.

192 BROADBAND tored by an individual (trustee) appointed to do so. This potential tem- plate for the rest of the industry was hammered out under significant pressures—above all the threat of the merger being blocked if the FTC’s concerns were not addressed. Meanwhile, in 2001, AT&T—prior to a wholesale revision of its business plan—appeared to be moving toward open access through its business strategy, influenced by regulatory and legislative scrutiny as well as a larger transition in business emphasis and strategy. 22 This change in strategy, ironically, was initiated by AOL (as an ISP) prior to its cable acquisition. Other cable operators may offer similar terms to forestall the long-term threat of federal intervention in the cable broadband business. In addition, overbuilders have begun constructing hybrid fiber coaxial cable systems in some markets that ISPs will be able to use to provide high-speed services. Whether this pattern of opening up will prove to be widespread and enduring remains to be seen. Already, similar activity has been stimu- lated in connection with another Internet technology, instant messaging, itself a factor in the government’s approval of the AOL Time Warner merger and an emphasis in the FCC investigation into the merger. The evolving open access situation, which illustrates a government role differ- ent from regulation per se, underscores the role of political activity in the shaping of telecommunications outcomes. It raises the prospect of ISPs making appeals to federal and state regulators to seek adjustments in access terms and prices and to antitrust authorities. Under these circum- stances, it is hard to ascertain what facilities operators or ISPs might do or might have done “voluntarily” as their judgments about business oppor- tunities evolved. Access Issues in Multidwelling Units Much of the debate about competition and open access assumed that broadband access is arranged at the sole discretion of the actual users. In the case of multidwelling units (MDUs), the establishment of facilities- based competition also depends on installation of new in-building facili- ties or working out some form of shared access to this infrastructure. MDUs, in which landlords may establish partnerships (including exclu- sive ones) with broadband providers, have become a new battleground 22For example, AT&T launched a recently concluded open access trial in Boulder, Colo- rado, at the end of October 2000 that connected eight ISPs: two AT&T affiliates (Excite@Home and WorldNet), two DSL providers (Winfire, Inc., and Flashcom, Inc.), two national ISPs (EarthLink, Inc., and Juno Online Services, Inc.), and two local ISPs (RMI.net, Inc., and FriendlyWorks, Inc.) (“ISP Offers ‘Open-Access’ Plan; AT&T Begins Trial in Boul- der,” Telecommunications Reports, November 6, 2000, p. 17).

BROADBAND POLICY AND REGULATION 193 for competitive and incumbent providers in both the residential and com- mercial markets. MDUs are an increasingly attractive and important early market opportunity in university and many urban communities, which are characterized by high-density housing, high density of broadband customers, or both. The committee did not explore the multidwelling unit issue at length, but notes that it is an area of ongoing debate over compe- tition and access, particularly because these typically high-density situa- tions are likely most attractive to would-be overbuilders. Landlord-provider arrangements can both improve the competitive- ness of real estate through improved telecommunications services and provide an additional revenue stream through business arrangements with providers. Not surprisingly, agreements often include exclusive ac- cess provisions, raising concerns on the part of both tenants and other providers about infringements on their access. Debate relates to access to buildings by competing service providers, the latitude of building owners vis-à-vis tenants to control providers’ access to their property and there- fore services available to tenants, and the role of federal and state regula- tors. CLECs and telecommunications equipment providers, for example, have formed “The Smart Buildings Policy Project”23 to address relevant issues via lobbying.24 This group appears to be at odds with a new class of providers that focuses on building access, BLECs (building-focused LECs), which may receive some investment by building owners.25 Meanwhile, building owners, with their own lobbying arms,26 argue that the issue falls under a real estate rather than a telecommunications regulation re- gime. Access and service issues also fall under the purview of local gov- ernments’ new municipal ordinances and cable franchising agreements. The whole area of access in MDUs awaits possible direction and clarifica- tion from the FCC and the courts. Access to Poles, Conduit, and Rights-of-Way Another dimension of access is access to utility poles and conduits. Facilities-based wireline entrants cannot enter a market if they cannot obtain access to the poles and conduits to install their facilities. Issues of concern include prices, terms of access, processing time for an order, and the rate at which access is provided. The situation is complex, with varied 23See <www.buildingconnections.org>. 24 “Lawmakers Ask FCC to Hold Off on Building Access Rulemaking,” Telecommunica- tions Reports, September 4, 2000, pp. 9-10. 25“Building Owners, Carriers Spar over FCC Proposal to Block Service, Extend Ban on Exclusive Pacts,” Telecommunications Reports, January 29, 2001, pp. 27-28. 26An example is the Real Access Alliance; see <www.realaccess.org>.

194 BROADBAND ownership arrangements (e.g., ownership by the electric utility or shared ownership between the electric utility and the ILEC); various cost-sharing arrangements among users; and circumstances in which different rates apply to ILECs, CLECs, and cable operators.27 The Federal Pole Attach- ment Act28 in effect requires utilities to provide cable companies with nondiscriminatory access to poles, ducts, conduits, and rights-of-way. It gives the FCC jurisdiction to ensure reasonable rates and access terms unless a state chooses to regulate these. Various state measures aim to increase access. Municipalities also have several options for using their control over rights-of-way to ease entry by new providers. EXPANDING ACCESS AND UNIVERSAL SERVICE POLICIES The extension of universal service policies to new infrastructure net- works has generally enjoyed public support, as seen in the cases of the Postal Service, electricity, and the interstate highway system, as well as telephony. Because broadband technologies and services will not be de- ployed everywhere at the same time and some areas will lag in service availability and/or performance, policy makers almost certainly will face claims from different sectors that they should intervene to ensure that broadband services are available in a shorter time frame throughout the country. It is important, therefore, to establish (1) where access is funda- mentally constrained without government intervention, (2) who is not likely to be reached after the broadest likely private sector deployment is achieved, and (3) how and when to intervene. Rationales for Intervention Generally speaking, competitive markets are assumed capable of de- livering the types and quantities of goods and services that consumers desire at the prices they are prepared to pay. In some cases, however, a competitive market may fall short of societal objectives, and governments may act to promote some kind of equity when the efficient outcomes associated with the market do not. A key premise of universal service programs has been the assumption that access to the service is essential for meaningful participation in the social, economic, and political sys- 27Peter Blum. 2001. Pole Attachment Rules: Establishing a State Policy. Slides from briefing to National Association of Regulatory Utility Commissioners’ Summer 2001 Telecom Com- mittee Meeting. Available online at <http://www.naruc.org/Committees/Telecom/ pole_attachment.ppt>. 28Section 224 of the Communications Act of 1934, as amended, 47 USC 224.

BROADBAND POLICY AND REGULATION 195 TABLE 5.1 Density and Demand Factors in Universal Access High Density Low Density High Demand Competition/market Universal service mechanisms, demand solutions adequate aggregation, new technologies Low Demand Economic and community development approaches Most challenging tems; drawing a similar conclusion with respect to broadband will, of course, involve assessment of how functionality is tied to broadband ser- vice and to what extent narrowband alternatives can substitute for broad- band in providing socially desired services. The price of a good or service deemed essential or socially important may limit its accessibility to significant groups of consumers. For example, the marketplace for food may be competitive, but government nonethe- less may decide to intervene in order to ensure that low-income consum- ers are able to buy enough to sustain themselves (e.g., through food stamps). Consideration of the rural challenge, and contrasting it to certain ur- ban environments, highlights the fact that lack of deployment may reflect either lack of demand or insufficient density for profits to motivate in- vestment. Concerns about access can be lumped, roughly, into two cat- egories: (1) that broadband services be available ubiquitously at some reasonable price and (2) that broadband services be affordable to most or all of the population. The appropriate interventions in a given location depend in part on the region’s density and demand characteristics (see Table 5.1). An affluent urban neighborhood might be characterized by high density and high demand, while a poor urban neighborhood might have high density and low demand. A typical rural area would have low demand and low density, while a rural community with an active tech- nology user base might have low density but high demand. Both of these concerns, together sometimes referred to as the “digital divide” problem, have received national and global attention.29 Access and use disparities are difficult to gauge, because Internet access is chang- ing comparatively quickly and because multiple, interdependent factors 29Global forums concerned with digital divide issues include the International Telecom- munications Union Development Forum, the European Union, and the G-8. U.S. founda- tions looking at digital divide issues include the Markle and the Benton Foundations.

196 BROADBAND (age, income, education, ethnicity, and so on) are associated with lower use rates, and geographical disparities in access overlay these other fac- tors. Density versus demand is just one possible decomposition of the problem. “Digital divide” means different things to different people and in different situations. There are many “divides.” On the technology supply side, the term may refer to unmet demand for high-capacity long-haul transport facilities, connections to the Internet backbone, as well as high- speed local access facilities (DSL, cable, or wireless). To some customers, digital divide means the time it takes to get a high-capacity T1 line in- stalled, or the price to use existing services. To a business customer with a business-to-business Web strategy, it means the need for redundant backup facilities in case one access path is interrupted, for example by a cable cut. Supply-side problems generally require approaches that en- courage building facilities, either directly through application of funds or indirectly by providing incentives or demonstrating or aggregating de- mand to attract provider investment. On the demand side, digital divide may refer to having less access to computers, to Internet connections, and to training or content if one is a rural resident, senior citizen, Native American or other minority, or in a family with lower income. Demand-side strategies tend to be best ad- dressed through a different set of strategies, including development of locally useful content that stimulates increased interest and applicability, creating community networking partnerships or providing public access points or local training in a school, library, community center, or “cyber cafe” that provides lower-cost alternatives to residential service. Such efforts may also have to face another sort of digital divide—depending on how such performance-enhancing elements as caches or content distribu- tion systems are implemented, nonprofit groups may be at a disadvan- tage compared with commercial producers that are able to pay more to ensure quality of access to their content. These considerations may be incorporated in economic development programs, although even experts in rural technology deployment note that “infotainment” may be the most important driver for demand.30 Of course, supply-side and demand-side problems can be related. Low density makes it more expensive to build facilities. Lack of demand makes it more difficult to recover the cost of building infrastructure. Eco- nomic development programs (generally at the state and local levels) naturally treat communications infrastructure, including broadband, as an element of local economic development. Being more holistic in nature, 30Andrew Cohill at June 2000 workshop of the Committee on Broadband Last Mile Tech- nology.

BROADBAND POLICY AND REGULATION 197 programs tend to combine cultivation of demand with provision of ac- cess, education, and opportunities; if successful, they increase willingness to pay. Implicit Transfer Mechanisms Used for Universal Telephone Service In contrast to programs for which support is explicit, such as the food stamps program, the support schemes in the telephone industry histori- cally relied on implicit mechanisms. The history of the telephone industry shows that government intervention has helped to overcome actual or perceived barriers of substantial cost differences among geographical ar- eas as well as to enable basic telephone service access among qualifying low-income individuals (e.g., through Lifeline and Linkup programs). For context, note that nationwide telephone subscribership is between 94 and 95 percent, but the penetration level for households with annual incomes under $5,000 is only 79 percent, as compared with almost 99 percent of households with incomes of at least $75,000 (a few of the high- income houses may have opted for wireless service instead).31 Over the years, although the goal of universal service remained a cornerstone of state, and to a lesser extent, federal regulation, the mean- ing of that term changed as the network evolved. In the 1920s and 1930s, universal service probably meant a single telephone in a geographic terri- tory. In the 1940s and 1950s, universal service may have meant access to party-line service. In the 1960s and 1970s, universal service may have meant access to single-line service. And, more recently, universal service has been interpreted to mean touch-tone service and access to more ad- vanced services. Stated simply, business long distance customers paid more for ser- vice and residential customers paid less for connections to the public switched network. Similarly, rural customers paid substantially less than did urban customers compared to the relative cost of their lines. One view of this situation, commonly held, is that in each case, one group is subsi- dizing the other. Other observers argue that it is natural that different types of customers will make different contributions to the common cost of the network.32 Whether one characterizes them as natural rate differen- 31“Phone Subscribership Holds Steady at 94.4%,” Telecommunications Reports, December 18, 2000, p. 21, summarizing the FCC report Telephone Subscribership in the United States, which is available online at <www.fcc.gov/ccb/stats>. 32See J.C. Panzar and S.S. Wildman. 1995. “Network Competition and the Provision of Universal Service,” Industrial and Corporate Change, vol. 4, no. 4, pp. 711-719. See also David Gabel. 1999. “Recovering Access Costs: The Debate,” in B. Cherry, S. Wildman, and A. Hammond IV, eds., Making Universal Service Policy: Enhancing the Process Through Multidisciplinary Evaluation. Lawrence Erlbaum Associates, Mahwah, N.J.

198 BROADBAND tials or implicit transfers, the following distinctions have been the subject of telephone policy and practice: • Business versus residential. Business customers typically are charged higher rates for local telephone service than residential customers, even though the cost of serving business customers frequently is lower (be- cause they are located in areas with higher population densities). These price differentials relative to other customers are justified on a variety of grounds, including the “positive externality” produced by maximizing the number of customers connected to the telephone network. That is, the value of being connected to the telephone network increases as more subscribers are added. Also, because the existence of a business can de- pend on communication, businesses typically value communications more than residences do (including the value businesses attach to calls to and from residential customers), and it can therefore make sense, even in competitive markets, to set prices that reflect such different valuations.33 • Long distance versus local service. Typically, the largest single cost of providing telephone service is the cost of the loop that connects a cus- tomer to the first point of switching in the telephone network, the “local loop,” the centerpiece of the last mile. The cost of a local loop is fixed (i.e., it does not vary with the number or duration of calls that are placed or received, up to its full capacity). Because the local loop is necessary to providing both long distance and local service (as well as any other tele- phone-delivered services), its costs are common to both. Economists are used to saying that there is no a priori appropriate way of allocating common costs among the different products that are jointly supplied by the associated assets.34 Historically, regulators required incumbent tele- phone companies to recover part of the local loop costs from the flat-rated 33While the elasticities of the different customer classes are not well understood, it is likely that this results in a situation where customers with the least elastic demands pay the highest price, which is the general relationship that one gets with Ramsey pricing. An interesting question is whether this could be a competitive outcome. Historical work by Gabel and a formal model by Panzar and Wildman suggest yes, though elasticity was not an issue in the model. Furthermore, more traditional models of competition allow for price discrimination. Baumol and Willig have argued in New Zealand regulatory proceedings that competition will necessarily generate Ramsey prices. (See Panzar and Wildman, “Net- work Competition and the Provision of Universal Service,” and Gabel, “Recovering Access Costs: The Debate,” 1999.) 34In most economic markets, the various products produced with common assets all make contributions to the common costs. Thus, for motion pictures, the fixed cost of pro- ducing a film is covered by earnings from theaters, videocassettes, pay television, and over- the-air broadcasting, not to mention foreign markets.

BROADBAND POLICY AND REGULATION 199 monthly charges assessed to end users for local service, part from the short-haul toll rates paid by end users, and the balance from per-minute access charges paid by carriers providing interstate and intrastate long distance service. The precise manner by which the common costs are allocated will have differentiated impacts on consumers. While it has the effect of holding down the price of telephone service for those who place few long distance calls, the practice of recovering part of the fixed cost of the local loop from usage-based prices, some observe, has been inefficient and has artificially depressed demand for long distance service. • Urban versus rural. The cost of providing local telephone service to urban customers is generally lower than the cost of serving rural custom- ers. The higher population densities enable a telephone company to serve a greater number of customers from a single switch, and the loops con- necting an end user to the first point of switching generally are shorter in urban areas. At the same time, this factor may be offset owing to the greater benefit obtained from technologies that overcome distance and dispersion, and there may be more willingness to pay by rural users, other things equal, given perceived value.35 Federal and state regulators traditionally have implemented complex cost and price averaging and other policies to maintain prices for basic telephone service in rural and other sparsely populated areas at levels comparable to and often even lower than those paid by subscribers in urban areas.36 State regulators historically required telephone companies to average their rates over large geographic areas so that customers in densely and sparsely populated areas paid the same rates.37 High-cost programs transfer funds for the purpose of providing service in low-density, expensive-to-serve areas. Because they support the upgrade of telephone facilities, high-cost funds can also indirectly contribute to increased DSL availability as well as increasing dial-up modem line speeds.38 35As noted by Richard Civille in his remarks before the committee in June 2000. 36Full parity is not the goal. For example, rural customers have much smaller local calling areas (the areas in which local calls are covered by the monthly flat rate) and as a result may pay much higher total bills. 37Some rural states adopted forms of rate deaveraging by, for example, requiring custom- ers in some areas to pay a “zone charge” in addition to the averaged, basic rate. 38An area of current debate is whether the high-cost fund should be explicitly expanded to cover broadband (and, a related question, whether caps on these funds should be relaxed to support advanced services build-out). Proponents of such changes argue that they are valuable mechanisms for enhancing rural infrastructure. For example, the Federal-State Joint Board on Universal Service’s Rural Task Force recommended that the FCC adopt a “no barriers to advanced services” policy that would permit high-cost funds to be used in ways supportive of providing advanced services, including reducing loop lengths, remov-

200 BROADBAND These support mechanisms have gradually been reformed in various ways following passage of the Telecommunications Act of 1996 (e.g., a shift from usage rates to flat rate recovery). But they still depend in part on one form or another of implicit transfer and still reflect assumptions about technology and the market more rooted in voice than in broadband services. The sustainability of implicit support schemes is in question when those services are not provided on an integrated basis by a regu- lated monopoly. Entrants will tend to target the “subsidizing” customers, since the new entrants are not required to offer service at “subsidized” rates and, consequently, the price they must offer the existing “subsidiz- ing” customers need only be lower than the incumbent’s price including the subsidy, all other things being equal. Similarly, entrants will tend to ignore the “subsidized” customers because the price they must offer those customers must be less than the incumbent’s price, including the subsidy. A very substantial part of the challenge that the FCC and state regulators face in opening local markets to competition is the need to reform the historical system of implicit support to make universal service support compatible with competition. Universal service is not inconsistent with the introduction of competition, but the historical scheme for maintaining universal service is inconsistent with competitive markets for local ser- vices. The issue of which firms should contribute (and benefit from) uni- versal service funds is, not surprisingly, a subject of ongoing political and regulatory debate. Despite these drawbacks, proponents of implicit sup- port mechanisms note that by virtue of the internal, largely unseen trans- fers, these mechanisms have the advantage that they may be less-politi- cized and more-stable sources of funding. Other Mechanisms for Increasing Access to Broadband Loans and Grants One avenue being pursued by governments, foundations, corpora- tions, and civic groups is partnering to leverage resources and carry out programs that expand access for underserved rural and urban popula- ing bridge taps, and otherwise upgrading the network to support DSL (Rural Task Force, Federal-State Joint Board on Universal Service, 2000, Rural Task Force Recommendation to the Federal-State Joint Board on Universal Service, submitted to the Federal Communications Com- mission under CC Docket 96-45, Sept. 29, available online at <http://www.wutc.wa.gov/ rtf/rtfpub.nsf>). Critics question whether the program should be expanded beyond tradi- tional telecommunications services, and the impact of any increased transfer of funds from low-cost to high-cost areas.

BROADBAND POLICY AND REGULATION 201 tions. Loans or matching grants could be considered for other instances of community-initiated efforts to develop local broadband networks. Programs of telecommunications regulatory agencies are comple- mented by other kinds of programs that may also contribute to connectiv- ity as part of a broader set of programs that support economic develop- ment and quality of life in underserved areas. The federal government has supported community access through the Department of Housing and Urban Development’s Neighborhood Networks program39 and the U.S. Department of Education’s Community Technology Center grants.40 The Rural Economic Development Act of 1990 (P.L. 101-624) created the Rural Development Administration within the U.S. Department of Agri- culture (USDA), which has other relevant entities and programs, and the Rural Electrification Administration makes loans and provides technical assistance to rural telecommunications providers. For example, the USDA’s Rural Utilities Service41 supports rural electricity, water, and tele- communications infrastructure through loans, grants, and technical guid- ance. The Rural Utilities Service launched a 2001 pilot program to provide $100 million in 10-year loans to companies building broadband infra- structure in rural areas. The program is targeted to communities with up to 20,000 residents, and following the FCC lead, uses a 200-kbps transmis- sion threshold to qualify for broadband status.42 Telephone cooperatives provide telephone service and a range of data services in a number of rural areas, and rural telephone companies have been active in deploying broadband services.43 Cooperatives can help aggregate demand across a widely distributed set of customers.44 There are also focused cooperative 39See <http://www.hud.gov/nnw/nnwindex.html>. 40See <http://www.ed.gov/offices/OVAE/CTC/factsheet.html>. 41According to 7 CFR §1735.10 (a), “The Rural Utilities Service (RUS) makes loans to furnish and improve telephone service in rural areas. Loans made or guaranteed by the Administrator of RUS will be made in conformance with the Rural Electrification Act of 1936 (RE Act), as amended (7 U.S.C. 901 et seq.), and 7 CFR chapter XVII. RUS provides borrowers specialized and technical accounting, engineering, and other managerial assis- tance in the construction and operation of their facilities when necessary to aid the develop- ment of rural telephone service and to protect loan security.” See <http://www.usda.gov/ rus>. 42“RUS Sets $100M for Rural Broadband Rollout,” Telecommunications Reports, December 11, 2000, p. 7. See also <http://www.usda.gov/rus/telecom/initiatives/initiatives.htm>. 43National Exchange Carrier Association (NECA). 2000. NECA Rural Broadband Cost Study: Summary of Results. NECA, Whippany, N.J. Available online at <http://www.neca.org/ broadban.asp>. 44Richard Civille, Michael Gurstein, and Kenneth Pigg. 2001. “Access to What? First Mile Issues for Rural Broadband,” white paper; see Appendix C. See also the National Tele- phone Cooperative Association <www.ntca.org>, which publishes Rural Telecommunications.

202 BROADBAND financing organizations.45 Finally, a variety of corporate initiatives have also provided support for community network access. The federal government has also intervened in the form of the e-rate program, originating in the Telecommunications Act of 1996 and launched in 1998, to provide funds (partial or matching support) for high-speed access for schools, libraries, and health care facilities.46 The e-rate pro- gram uses funds raised from taxes levied on particular communications services to expand public access broadband services through these facili- ties. The program is associated with significant increases in school con- nectivity, where its influence has been to accelerate, enhance or comple- ment, or enable that connectivity, and there is evidence that it is supporting increases in the bandwidth of connections sought by schools (e.g., movement from “plain old telephone service” to T1 and then T3 lines).47 Local points of presence and applications have also been supported through an evolving effort that has been more like a demonstration pro- gram—the Technology Opportunities Program (TOP) (originally the Tele- communications and Information Infrastructure Assistance Program)—at the U.S. Department of Commerce’s National Telecommunications and Information Administration (NTIA). The program began in 1994 and fo- cuses on model uses in public sector and nonprofit contexts. Whereas some $2 billion have been awarded under e-rate, TOP awards total about $150 million.48 Collectively, these programs increase awareness, foster deployment and use, and extend broadband beyond those consumers most likely to go after it on their own. Tax Incentives Another option for promoting access is to provide tax incentives for investment in high-cost areas. Tax options to promote broadband deploy- ment in high-cost and/or low-income areas are appealing on two grounds. First, they avoid long-recognized economic distortions associated with funding universal service goals through revenues raised by taxing certain services within the specific industry (the approach used for basic tele- 45See National Rural Utilities Cooperative Finance Corporation (<www.nrucfc.org>) and its affiliates. 46E-rate involves discounts on eligible facilities’ purchases of telecommunications and Internet services plus internal networking, with discounts varying with location (e.g., high- cost, low-income). 47“Telcos, System Integrators See Rising ‘E-rate’ Revenues,” Telecommunications Reports, September 11, 2000, pp. 46-48. “Poorer, Larger Applicants Get More ‘E-rate’ Funds, Study Finds,” Telecommunications Reports, September 18, 2000, p. 29. 48See <http://www.ntia.doc.gov/otiahome/top/grants/briefhistory_gf.htm>.

BROADBAND POLICY AND REGULATION 203 phone services, and now, through the e-rate, for broadband connections to schools and libraries). Such levies applied selectively within an indus- try can distort relative prices and, thereby, choices made by the users of communications services.49 Selective, within-industry levies of this type also may create financial incentives for investments in economically inef- ficient facilities and services that bypass the providers and services sub- ject to the internal tax. Thus, ILECs have frequently argued that many of the investments in network facilities by CLECs were motivated by the financial payoff to be realized by providing long distance carriers and local businesses with the opportunity to avoid local access charges; simi- lar thinking for plans to use cable plant to offer local service was also one of the reasons AT&T offered for its purchase of cable operator Tele-Com- munications, Inc. (TCI). Investments motivated by the avoidance of extra charges created by policy interventions rather than the pursuit of produc- tion cost-driven competitive advantage are inefficient and a waste of soci- etal resources. Tax credits (and other subsidies such as the high-cost fund support mechanism) still may provide incentives for what is ultimately uneconomic activity. Second, reliance on tax credits to finance broadband deployment would mean that the federal government would have to consider the financing of communications policy goals in the context of the larger set of societal trade-offs that necessarily must be addressed in setting and allocating federal budgets. To the extent that tax credits are used, they would put the financing of communications policy goals squarely within the traditional budgetary process. This might force a better integration of communications policy goals into the larger set of societal goals addressed through various types of federal funding, so that the relative merits of communications policy goals might be appropriately assessed in com- parison with other social policy goals in the allocation of scarce govern- ment resources. Because telecommunications spending would compete against many other interests, it might also mean less-stable funding for telecommunications programs. Vouchers Another way of putting financing “on the books” is to issue vouchers, similar to food stamps. As a budget line item, a voucher program is even 49For example, contributions to the universal service fund have traditionally been built into charges that long-distance carriers pay local exchange carriers for completing their calls over local networks. Because such policy-driven charges must necessarily be recov- ered in the price of long-distance calls, the price of long distance increased relative to the price of local service and other communications services.

204 BROADBAND more evident in the decision-making process. Broadband-specific vouch- ers might be a useful tool for promoting broadband penetration. It is important, however, that such vouchers be targeted as narrowly as pos- sible to specific groups of consumers (e.g., low-income consumers or those living in high-cost service areas) who are not likely to subscribe in the absence of such aid. There is little point in subsidizing purchases that would be made without the subsidy. Vouchers are especially attractive in situations in which infrastructure deployment is not an issue (service providers already have an incentive to build out an area), but to which some subset of potential customers in a built-out service area would not subscribe at prices that service providers would have to charge to cover their costs.50 In these circumstances, a voucher can be a highly specific instrument for encouraging subscriptions that would not happen other- wise. It is less clear that vouchers have advantages over direct payments to service providers if the goal is to promote infrastructure deployment in areas that might not otherwise be served. To the extent that providers might compete to offer services to customers in such areas, there is the danger that competitive lowering of service prices would transfer some portion of the voucher to consumers’ pocketbooks rather than to covering infrastructure expenses. A more efficient approach in these areas might be to let providers bid for the right to serve these territories. Research to Develop Technology Alternatives Finally, there is the option of supporting research as a means of pro- moting access. Much of the excitement associated with progress in broad- band technologies and the diffusion of fiber builds on research and devel- opment that has enabled new technological approaches and/or lowered the costs or increased the performance of existing approaches. The federal government is key in supporting basic research and fostering public dis- semination of the results of research. Units of the U.S. Department of Defense (notably the Defense Advanced Research Projects Agency, or DARPA), the National Science Foundation (NSF), and other federal orga- nizations are key supporters of communications R&D. Most of that R&D recently has focused more on technologies and components that enhance network backbones or development of applications that run over high- speed networks than on local access networks. But DARPA has had a program aimed at fiber in the distribution network, and NSF and DARPA have supported a variety of wireless networking research. 50Note that, from a life-cycle or total cost perspective, decreases in the cost of equipment associated with use of broadband, “CPE,” will also affect willingness to pay for service.

BROADBAND POLICY AND REGULATION 205 Looking Forward For policy makers, the threshold issue is how to determine whether government intervention to accelerate broadband deployment is neces- sary or desirable. It appears that the problem is not whether most areas will ultimately have some form of broadband service, but rather that in rural areas deployment will occur well after such services are available in more densely populated areas or that the technology options and/or per- formance will be different in rural areas. The leading broadband technologies today (in terms of installed base and technology maturity) are both wire-based, and it seems likely that for the near term at least, distance and population density will deter their rapid deployment in remote or sparsely settled areas. Because of the added per-passing cost of serving rural areas, different kinds of technical strategies may need to be sought there as compared with those for other (denser) areas; an example would be greater emphasis on wireless links from residences to a fiber backbone (possibly leveraging local govern- ment or electric and water utility rights-of-way).51 With broadband satel- lite services—which may be able to serve these areas more cost-effectively than the wireline alternatives could—having recently been introduced to the market, one finds a situation where there is some form of broadband available in even the most remote areas of the continental United States. It is also encouraging that the initial offering prices of the Starband service suggest that the “rural penalty” may be small (recurring charges for satel- lite service at $60 per month versus the $30 to $50 per month typical of cable or DSL). However, it is unclear at this point whether these services will be able to achieve and maintain sufficient performance levels to serve as adequate substitutes for the functionality of wireline services, or how their cost and price will compare in the long run with wireline service in more densely populated areas. At today’s broadband penetration levels, it seems premature to make conclusions about the shape of deployment. Consumer technologies gen- erally display an S-shaped adoption curve, which is marked by an initial period of slow adoption, followed by more rapid expansion, and, finally, a leveling off of adoption in the later stages. In the case of narrowband Internet access, NTIA data collected over the past half-decade show that overall access has expanded greatly and that some disparities—such as across sex and race/ethnicity—have narrowed over time, primarily through expansion of dial-up household access and access in the work- place or public facilities. However, this access has largely leveraged near- 51See Civille et al., “Access to What?,” 2001.

206 BROADBAND ubiquitous public telephone network lines, and thus, with the exception of some instances where line quality is very poor, has not been hampered nearly as much by technological and economic constraints on where and when new facilities are deployed as broadband would be. Widespread dial-up use suggests that wide segments of the population find Internet access to be of value, and thus suggests widespread demand for broad- band. In the case of residential broadband, deployment has been growing rapidly from a presently small base, from which vantage point it is hard to infer the long-term adoption rate, patterns of availability, or the ulti- mate level of adoption. To the extent that policy makers are simply uncertain about the pace of broadband deployment, the benefits of government intervention to accelerate that process would have to be clear and substantial in light of the risk that such intervention may have unintended and undesirable consequences. Although government policies likely contributed to the high penetration of telephone service in rural areas, application of such policies to broadband could, in theory, deter future entry by competing broadband providers that cannot match the below-cost rates resulting from averaging and other distributional policies. Another risk is that by picking particular technologies or defining particular services, some gov- ernment programs aimed at bringing a technology to all may end up freezing the technology deployed. Policy makers seeking to promote rapid, efficient broadband deployment should assess the effectiveness of strategies that help avoid these risks—including demand stimulation and aggregation, grant and loan programs, and municipal initiatives fostering market entry and competition. This analysis would require policy makers to collect and review reliable broadband data on an ongoing and timely basis. The development of a comprehensive, national universal service program may well become desirable in the future, once the pace and scope of broadband deployment become clearer. THE LOCAL ROLE IN BROADBAND Seeking to accelerate or enhance the delivery of telecommunications services in their communities, a number of cities, counties, and states have considered or launched initiatives aimed at facilitating, encourag- ing, or directly building infrastructure for broadband. Historically, the direct local role with respect to telecommunications has been limited largely to negotiating cable franchises. In addition, local governments— absent preemption from higher levels—have control over local features of the deployment environment, such as public rights-of-way, zoning, per- mitting, and so on. These practical issues affect decisions about special facilities, such as “carrier hotels” and data centers. Local government

BROADBAND POLICY AND REGULATION 207 influence has been expressed in conflicts over siting for terrestrial wire- less towers and satellite dishes. Local governments also control access to rights-of-way and proposals for local investment in conduit that can be deployed once and that contain cable or wire supporting multiple provid- ers and services. Today, communities are exploring how to use these points of leverage as well as other mechanisms and incentives to promote broadband deployment. Local governments have a direct interest in neighborhood, commu- nity, municipal, and regional infrastructure, and it is within the commu- nity that existing government, corporate, university, and school networks are deployed. Local governments may be in a better position than na- tional providers are to collect and verify local marketplace information, such as discovering and/or aggregating latent demand for broadband services, and these governments involve people whose jobs involve satis- fying local interests. Indeed, where local entities have moved to provide local infrastructure, it has typically been when no commercial firm was willing to invest in a given community. Local initiatives are not without their critics, however. For instance, while local decision makers may see benefits from broadband, it can be as hard for them as for service providers to predict and elicit consumer demand52 and design sustainable business models for municipal broad- band enterprises. Telephone and cable incumbents tend to protest local efforts to serve more than government users with services procured or provided by government entities.53 Critics also argue that locally based efforts are less likely to be commercially sustainable in the long run, suf- fering regularly from lack of access to capital to support upgrades. Local efforts may also have insufficient economies of scale to be viable in the long run, and risk becoming overly politicized. Local and regional broadband initiatives cover a wide range of possi- bilities, from focusing on local government infrastructure to facilitating access for the community at large. Local approaches vary for obvious reasons—size, local market desirability, and whether existing providers have been introducing broadband service. This variation may ultimately 52Civille et al. (“Access to What?,” 2001) argue for demand cultivation in combination with access promotion through community economic development programs. Their ac- knowledgment that growing demand may take work—that simple access is not sufficient— implicitly supports the view that accelerating deployment is risky. 53”Private-sector carriers say they shouldn’t have to compete with the entities that regu- late their rates, grant them operating certificates and franchises, control their access to vital rights-of-way, and tax them” (“Community Size: The Difference in Cities’ Telecom Choices?” Telecommunications Reports, December, 4, 2000, pp. 36-38).

208 BROADBAND limit what can be learned or replicated from any specific instance, but an informal network of supporters of local efforts has fostered the exchange of relevant information, including approaches to architecture, contract- ing, and financing, to maximize opportunities for local officials to learn from others.54 The examples listed in Box 5.2 indicate the sorts of initiatives that have been undertaken at the local or regional level.55 They include, for example, public operation of a multiservice network, where a municipal or county agency is the operator, providing retail services to the end user. Municipal service monopolies are not unusual—water or sewer authori- ties are the dominant model, and public power utilities are found in a number of locales—but the high level of complexity and rate of change in telecommunications technology compared with water or electricity sup- ply pose a risk. And as noted above, this approach may raise objections from private sector providers, who see the public service as being unfairly subsidized. However, if there is no private sector provider on the horizon, it may be an attractive option. Another option is some form of public-private partnership. Again, this raises concerns about the risk in a government body’s entering into what may be a long-term relationship with a selected private sector player. If the relationship sours, it may be difficult to replace the private sector player. This sort of approach runs the same risk as that with exclusive cable franchising—communities may derive revenue or other benefits from the arrangement, but the partner may not deliver the level or quality of service desired. Additional complications arise if the public sector has contributed funding to the venture. The drawbacks of the approaches listed above argue that local gov- ernments concentrate on taking steps to encourage and facilitate com- petition among private sector players rather than creating new quasi- monopolistic entities. As a public sector partner with multiple private providers, a public agency would not be competing with a private sector retail service provider. Another advantage of this strategy is that it means that private sector providers do not have to negotiate with each other to 54One ongoing effort that attracted the attention of this committee has been the work of Bill St. Arnaud as part of Canada’s CANARIE program. See <http://www.canarie.ca/>. 55The Community Broadband Deployment Database, established by the National Regu- latory Research Institute at Ohio State University for the Federal Communications Commis- sion, lists more than 200 community broadband programs, covering a range of technolo- gies, target user groups, and funding sources. See <http://www.nrri.ohio-state.edu/ programs/telcom/broadbandquery.php>.

BROADBAND POLICY AND REGULATION 209 obtain access to facilities, which reduces the need to regulate their con- duct. Easing access to rights-of-way is the simplest step, but this may not be enough to induce new entrants. Another option is for a local or regional government agency to install fibers (or conduits through which fibers can later be pulled) and use this investment to lower the barriers to entry by private sector players by making the infrastructure available to them. This approach can be implemented in a number of ways. One is the “fiber condominium” model, in which a locality declares its intention to build out fiber along its streets and invites any interested parties to purchase some share of the fibers installed (and possibly installs additional dark fiber for future use). Typically, some provision is made to lease colocation space for service providers at the fiber termination points. Alternatively, the locality may enter into partnerships with one or more private sector companies to install (and possibly maintain) the fiber. The town itself can sign up, as can schools and municipal departments, businesses and other private sector players in the town, citizens themselves, and any interested broadband providers. The locality provides the motivation and coordina- tion for joint action—it shares in the cost of the common construction, but it may also prohibit the digging up of streets again for some period after the construction. By avoiding the extra cost of uncoordinated overbuild- ing—keeping down the per-passing costs—this approach attempts to pro- vide competition at per-passing costs comparable with those of a single provider. Local and regional government or quasi-governmental agen- cies can also act in effect as anchor tenants that underwrite some of the cost of installing infrastructure, reducing the costs for other government agencies, private sector firms, or even individual customers. The conse- quence of this action is that more providers may be motivated to enter the market in the town. Finally, localities may choose to launch experimental pilot projects to explore new technologies, system architectures, or business models. State or federal grants can help support communities exploring options or en- able them to purchase facilities that today are more costly than they will be in the future when suppliers are able to achieve scale economies in the production of such equipment. Such pilot efforts can demonstrate the viability of systems, demonstrate the extent of demand for them at the local level, and support achievement of scale in use, either by closing access gaps or increasing interest in use. The state or federal role is appro- priate, given that results of the experiment can help inform future private sector or public sector initiatives using similar systems. New develop- ments can also build in broadband infrastructure, as is beginning to hap- pen (Box 5.3).

210 BROADBAND BOX 5.2 Some Examples of Municipal and Regional Broadband Initiatives • Blacksburg, Virginia. The Blacksburg Electronic Village (BEV) initiative began in 1991 with a public-private agreement between the town of Blacksburg, Bell Atlantic (now Verizon), and Virginia Polytechnic Institute and State University. When service launched in 1993, BEV provided Blacksburg residents with dial-up access. Also, starting in 1994, integrated services digital network (ISDN) and Ethernet have been made available to an increasing number of townhomes and apartments. Because dial-up access became avail- able from commercial providers, BEV turned over its modem pool customers to the private sector in 1995, and similarly transferred its Ethernet operations to the private sector in 1998. At present, 87 percent of Blacksburg residents are online, according to BEV infor- mation. Currently, efforts are under way to develop a townwide all-fiber network, to inte- grate wireless with wireline services, and to develop a broadband switch point and ex- change for advanced network services. BEV was prevented from extending service to surrounding areas because, after industry lobbying, a law was passed to preclude this government service provider role.1 • Abingdon, Virginia. In December 1995, a group of citizens met to discuss the potential of providing residents of Abingdon with high-speed Internet connectivity. These activities led ‘to the launch of the Electronic Village Abingdon (EVA) initiative. In addition to modem and ISDN connections, a fiber-optic connection from the town manager’s of- fice to the hospital and the Washington County Main Library was established as a partner- ship between the Town of Abingdon and Sprint. On the basis of results of early trials, the project was expanded. During Phase I, EVA’s fiber-optic service was extended in the downtown area, providing high-speed connections to every building within 150 feet of the fiber backbone. Phase II, under way in 2001, is a collaborative effort with Highlands Union Bank to extend the fiber-optic cable toward the west of Abingdon. The network provides 10-Mbps connectivity within the town network and Internet connectivity at 1.54 Mbps. Subscribers need to purchase a fiber-optic transceiver (about $150) and pay a one- time installation charge of $75. Monthly access fees are $35 per month for 10 Mbps and a single Internet Protocol (IP) address. Recently, 100-Mbps service has been added for $70 per month.2 • Berkshire County, Massachusetts. Berkshire Connect was established through a 1997 grant of $250,000 from the Berkshire legislative delegation and the Berkshire Re- gional Planning Commission. A project task force committee, with representatives from cultural institutions, local businesses, public access organizations, and local business con- sultants, was established to propose a strategy for enhancing Berkshire County’s infrastruc- ture and to reduce the cost of networking government agencies. The original strategy called for Berkshire Connect to partner with the private sector and use a mix of public and private funds to build the infrastructure, but Global Crossing/Equal Access, the winning bidder, agreed to build the infrastructure without using public funds. The agreement estab- lishing the service offers volume pricing, with prices for all members decreasing as more subscribers sign up. Berkshire Connect competes with the incumbent LEC, Verizon, in 1Blacksburg Electronic Village (BEV). 2001. “About the BEV.” Available online at <http://www.bev.net/ project/brochures/about.html#2>. 2See the Electronic Village Abingdon home page online at <http://www.eva.org/>.

BROADBAND POLICY AND REGULATION 211 both the retail and wholesale markets, though it claims more success in the wholesale market (providing backbone services to other providers). Berkshire Connect does not offer residential or last mile service at present, but is currently working to define a business strategy for last mile service. It is also exploring new business relationships with other Massachusetts regional networks.3 • LaGrange, Georgia. Through the LaGrange Internet TV initiative, the City of LaGrange, Georgia, provides Web access to all cable television subscribers in this com- munity of 27,000 residents. In 1998, Charter Communications, Inc., and LaGrange en- tered into a leaseback agreement in which the city financed and constructed a two-way hybrid fiber coax network (using city funds, without state or federal support). Using World- Gate’s Internet on EVERY TV Service, it allows LaGrange cable TV subscribers to have Web access at no additional cost.4 • Glasgow, Kentucky. In 1994, as Glasgow was being wired for cable, town officials decided to facilitate Internet traffic via cable as well. The 12-year-old Glasgow fiber-optic system, one of the first of its kind, provides relatively inexpensive cable and high-speed Internet services to 8,000 homes and businesses, or two-thirds of the local market for cable. According to its managers, it has broken even for the past 4 years. Proponents note that the network has been moderately successful in spurring local economic activity. For example, Franchino Mold & Engineering Co. is cited as having opened a new facility in Glasgow in 1998, in part because the city’s network allowed for an easy exchange of data with engineers at the company’s Lansing, Michigan, headquarters. Two-thirds of Glas- gow’s businesses and a quarter of its residences now pay for broadband cable. The town’s network connectivity is also used for a variety of other functions including controlling traffic lights, coordinating utility repairs, and plotting school bus routes.5 • Washington County, Ohio. Seeing inadequate broadband facilities in the county, the nonprofit Washington County Community Improvement Corporation (CIC) launched a nonprofit corporation, Sequelle, to provide terrestrial wireless broadband communica- tions to southeastern Ohio and the mid-Ohio Valley region, targeting business and educa- tional customers. Using a mixture of state and federal startup funds, it plans to launch service in 2001.6 • Chicago, Illinois. Chicago’s CivicNet is a citywide initiative to build a new broad- band infrastructure for government, businesses, other institutions, and residents. The city plans to use the $32 million it spends each year on voice and data communications to become an “anchor tenant” for a high-speed fiber-optic network, to be constructed in partnership by one or more lead technology vendors. The city also plans to make city- owned or -controlled conduits and tunnels available to reduce installation costs. The continues 3Sources include the Berkshire Connect home page, which is available online at <http:// www.bconnect.org/>; “FCC Hearing, May 22, 2000,” which is available online at <http:// www.bconnect.org/FCChearing5_00.htm>; and a personal communication with Bill Ennen, Donahue Institute of the University of Massachusetts, April 19, 2001. 4See <http://www.lagrange-ga.org>. 5See David Armstrong and Dennis K. Berman, 2001, “Municipal Networks Become Rivals for Fiber- optic Telecom Companies: Dissatisfaction Spurs Competition,” CNBC and The Wall Street Journal, August 17. Available online at <http://www.msnbc.com/news/615215.asp#BODY>. 6For more information, see the Sequelle home page online at <http://www.sequelle.com/>.

212 BROADBAND ultimate goal is for fiber networks to extend to every neighborhood, and ultimately down every street.7 • Lane and Klamath Counties, Oregon. The Lane Klamath Regional Fiber Consor- tium, formed by Lane and Klamath Counties and the cities of Coburg, Chiloquin, Klamath Falls, Lowell, Merrill, Oakridge, Springfield, and Westfir, negotiated joint agreements with Pacific Fiberlink (now Worldwide Fiber). In return for assistance with permitting and an exchange in lieu of right-of-way fees, the local governments received 12 strands of fiber installed in a contiguous strand extending approximately 200 miles from Coburg to Mer- rill, with points of access in the cities and significant county points along the route. Con- sortium members plan to use this fiber to increase communication opportunities to all the residents along the route and in communities adjacent to it.8 • Marietta, Georgia. In 1996, Marietta FiberNet became Georgia’s first municipally owned company to be certified as a competitive local communications carrier. Wholly owned and operated by Marietta, it is structured as a separate business. The company began constructing its network in the spring of 1997. Today Marietta FiberNet provides high-speed voice and data services to local schools, hospitals, and businesses over 170 miles of fiber-optic cables. • Thomasville, Georgia. Thomasville’s Community Networks Services (CNS) oper- ates a citywide fiber-optic network capable of supporting high-speed Internet access, ca- ble television, and energy management services, as well as a dial-up Internet access ser- vice. CNS now has more than 3,000 customers for its Rose.Net Internet service and expects many of those customers to move up to Rose.Net express, a broadband Internet service, when it becomes available. Thomasville recently entered into an agreement with neighboring Tifton through which it will help launch Tifton’s proposed Friendly City Net- work’s Internet service.9 • Grant County, Washington. The Grant County Public Utility District (PUD) initially launched its fiber network, Zipp, in order to upgrade its electrical substation control. It later decided that the network could also be used to provide telecommunications services in the county. Through Zipp, independent service providers can provide data, voice, and video services to their customers. The PUD’s fiber backbone was completed in 2000, and pilot projects aimed at commercial and residential customers are underway. In 2001, the PUD began marketing to the general public. Major residential fiber construction and build- out are slated to take place between 2001 and 2006 to reach 90,000 homes.10 • Muscatine, Iowa. Muscatine Power and Water (MP&W), the town’s incumbent, municipally owned public utility, was first to deploy high-speed facilities in Muscatine. Following a 1996 marketing study and detailed feasibility study, the communications util- ity was launched following approval by public referendum in 1997. The communications utility received $18 million in initial funding from the municipal electric utility and 7See Department of General Services, Bureau of Telecommunications and Information Technology, City of Chicago. 2000. “Request for Information: Chicago CivicNet” (Specification No. B09189503), November. Available online at <http://www.cityofchicago.org/CivicNet/civicnetRFI.pdf>. See also Tom Kontzer. 2001. “Chicago’s CivicNet Takes a Step Closer to Reality,” InformationWeek.com [January 4]. Available online at <http://www.informationweek.com/story/IWK20010104S0007>. 8See <http://www.ruralfiber.net/lkpage.html>. 9See Georgia Municipal Association. 1999. “Tifton and Thomasville Enter Internet Agreement,” Octo- ber 7. Available online at <http://www.gmanet.com/news/1999/1007.internet.shtml>. See also Tho- masville Utilities Community Network Services (CNS). 1999. “Press Releases.” Available online at <http://www.tucns.com/press.html>. 10See the Zipp Fiber-optic Network home page online at <http://www.gcpud.org/zipp/>.

BROADBAND POLICY AND REGULATION 213 completed construction of its fiber network in the spring of 1999. MP&W provides high- speed cable modem Internet access to residential customers and a Municipal Area Net- work for business customers. MP&W’s telecommunications network consists of a hybrid fiber coax system with 125 homes per node, which can deliver a maximum of 4 Mbps downstream and 1 Mbps upstream for connected customers. MP&W obtains its connec- tion to the Internet backbone through a division of Iowa Network Services, Inc. (INS), a telecommunications firm formed by a consortium of 128 independent telephone compa- nies.11 • Ashland, Oregon. Building on Ashland’s earlier fiber network initiative, local Inter- net companies are cooperating with the city to establish a new service, dubbed “Ashland Unwired.” In early 2001, the effort began with a demonstration at Ashland’s Starbucks coffee shop. Using access points running the IEEE 802.11b-standard wireless local area network technology, it aims to provide wireless Internet access via the city’s fiber network. Project A and Open Door Networks, city-certified Internet service providers, have offered to provide connectivity through the Ashland Fiber Network and assistance to help any business or organization wishing to provide Ashland Unwired to their customers.12 • Tacoma, Washington. The city’s public utility, Tacoma Power, began its network- ing activities with the construction of a fiber-optic network in 1997. In 1998, it launched cable television service as a competitor to the existing franchisee (TCI, now AT&T). In 1998, it began providing Internet service over the cable television network and began full cable modem service in 1999.13 • San Diego County, California. The High Performance Wireless Research and Edu- cation Network (HPWREN) project was launched in 2000 to build, demonstrate, and evaluate a noncommercial, prototype, high-performance, wide-area, wireless network in San Diego County. Built by researchers at the University of California at San Diego under a $2.3 million grant from the National Science Foundation, the network includes back- bone nodes on the UC San Diego campus and a number of rural areas in San Diego County, including the Pala and La Jolla tribes in remote San Diego County. Among HPWREN’s goals are to explore how scientists can make use of the network for real-time data collection and how the network can be used by rural Native American communities for interactive computer classes and remote tutoring programs. In addition to the research and education applications, the HPWREN team is also investigating ad hoc advanced network development and experimentation in collaboration with local crisis management agencies.14 11See <http://www.mpw.org>. Case study reported in Federal Communications Commission. 2000. “Inquiry Concerning the Deployment of Advanced Telecommunications Capability to All Americans in a Reasonable and Timely Fashion, and Possible Steps to Accelerate Such Deployment Pursuant to Section 706 of the Telecommunications Act of 1996: Second Report,” CC Docket No. 98-146, August. Available online at <http://www.fcc.gov/Bureaus/Common_Carrier/Orders/2000/fcc00290.pdf>. 12See Ashland Fiber Network. [Undated]. “Frequently Asked Questions.” Available online at <http:// www.ashlandfiber.net/index.asp?page=FAQ#1>. See also Ashland Unwired. 2001. “Open Door Net- works and Project A Unveil Wireless Internet Access in Ashland” [press release], January 29. Available online at <http://www.ashlandunwired.com/news.htm>. 13See Tacoma Power, Click! Network. [Undated.] “Project History.” Available online at <http:// www.click-network.com/news/history.htm>. 14See the HPWREN home page at <http://hpwren.ucsd.edu/>, as well as Hans-Werner Braun’s testi- mony before the House Science Committee’s Subcommittee on Research, July 31, Washington, D.C., available online at <http://www.house.gov/science/research/jul31/braun.htm>. See also “Wireless In- ternet to Native American Reservations,” PopularTechnologies.com, [undated], available online at <http://www.populartechnologies.com/news/01/02/15/0414200.shtml>.

214 BROADBAND BOX 5.3 Examples of Greenfield Developments New community and subdivision developments are increasingly incorporating broadband fiber-to-the-curb or fiber-to-the-home with other advanced exterior and interior networking infrastructure and services. These projects are often joint ven- tures between building developers who own the properties and rights-of-way; con- vergent telecommunications providers; and real estate investors and homeowners. Several examples follow: • Centennial, Indiana. HFC Internet, phone, and cable provided to 900 homes north of Indianapolis by the builder, Estridge Company, and First Mile Technologies. The cost will be repaid through homeowners’ association fees. • DC Ranch, Scottsdale, Arizona. All homes are being built with structured wir- ing to support in-home networking. • Celebration, Florida. In a Disney-built new community near Orlando, AT&T, Sprint, and Jones Communications have formed a joint venture to provide connectiv- ity.1 • Summerlin, Nevada. In a community outside Las Vegas, fiber-to-the-curb is being installed by the Howard Hughes Corporation and Sprint.2 • Valencia and Newhall Ranch, California. In communities developed by Newhall Land & Farming, SBC Communications is providing broadband voice, video, and data to homes through a revenue-sharing venture with the developer. • Hatchet Ranch, Colorado. Rye Telephone of Colorado City is installing fiber to 500 homes in an 80-square-kilometer new community.3 1See the Celebration, Florida, home page online at <http://www.celebrationfl.com/>. 2For more information, see “Summerlin: A Profile,” available online at <http://www. therousecompany.com/whoweare/hughes/summprofile.html>. 3Information from <http://www.fone.net.soco/guide/colocity/rtc/home.html>; Robert Pease. 2000, “Rural Areas Present Better Business Case for Fiber to the Home,” Lightwave, June, available online at <http://lw.pennnet.com/Articles/Article_Display.cfm?Section=Archives& Subsection=Display&ARTICLE_ID=73404&KEYWORD=Hatchet%20Ranch>; and Canet-3- NEWS, 2000, “Rural Areas Better Business Case for Fiber to the Home,” November 2, avail- able online at <http://www.canarie.ca/MLISTS/news2000/0185.html>. The risk in all of these possibilities is that the local government will not be equipped with the knowledge or skills to negotiate with a large private sector provider. If the town does not act carefully, there is risk of industry capture, an outcome in which a private sector provider manipu- lates the situation to the point where the town becomes dependent on it and thus loses any power to negotiate or foster competition. With some notable exceptions, local governments are less likely to be familiar with the technology and business side of networking than they are with more

BROADBAND POLICY AND REGULATION 215 traditional government operations, which places them at a disadvantage in planning or acquiring networking infrastructure or services than a private sector firm would be. The risk can be minimized if the town sticks to a facilitating role at the infrastructure level and encourages competi- tion from the outset. Still, industry is not monolithic, and some companies can be expected to favor and others to resist local efforts to foster market entry. Local governments, especially in smaller communities, often have limited capabilities. Action at the higher levels of government is an im- portant part of this local approach as well, to coordinate experiences, to catalog best practices, and to define the playing field with overarching regulation that prevents the obvious forms of mutual abuse.

Next: Bibliography »
Broadband: Bringing Home the Bits Get This Book
×
Buy Paperback | $48.00 Buy Ebook | $38.99
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Broadband communication expands our opportunities for entertainment, e-commerce and work at home, health care, education, and even e-government. It can make the Internet more useful to more people. But it all hinges on higher capacity in the "first mile" or "last mile" that connects the user to the larger communications network. That connection is often adequate for large organizations such as universities or corporations, but enhanced connections to homes are needed to reap the full social and economic promise.

Broadband: Bringing Home the Bits provides a contemporary snapshot of technologies, strategies, and policies for improving our communications and information infrastructure. It explores the potential benefits of broadband, existing and projected demand, progress and failures in deployment, competition in the broadband industry, and costs and who pays them. Explanations of broadband's alphabet soup – HFC, DSL, FTTH, and all the rest – are included as well. The report's finding and recommendations address regulation, the roles of communities, needed research, and other aspects, including implications for the Telecommunications Act of 1996.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!