Funding the National Information Infrastructure: Advertising,
Subscription, and Usage Charges
The stunning changes in information and communications technologies are focusing attention on the gap between the potential services available from new telecommunications networks and the much more limited services now available from our existing communications infrastructure. While this focus is useful in galvanizing political support for action, it may divert attention from the magnitude of the economic task involved in building the new information superhighway. Investors will be committing enormous resources to risky new technologies with uncertain information on future service revenues, the cost of these services, and the prospect of further technical breakthroughs that could render new facilities obsolete in a short time.
In this environment, policymakers must be careful not to foreclose the opportunity to develop new facilities, new services, and new sources of economic support for both. The federal and state governments should be encouraged to allow new services to develop over a variety of distribution systems and to permit vendors of these services to select the funding mechanisms for such systems or services. Indeed, it is precisely because we have already begun to open our telecommunications markets to competition and to reduce government regulation of them that we are now likely to develop the new national information infrastructure more rapidly than most countries and that other countries are moving to emulate us.1
In this paper, I review the recent historical data on the evolution of U.S. communications and media industries with a particular focus on the sources of revenue that have propelled this growth and the effects of restrictive government regulation. In addition, I provide a brief review of the potential costs of the new technologies and the uncertainties involved in implementing them. My conclusions are as follows:
NOTE: Robert Crandall is a senior fellow at the Brookings Institution. The views expressed herein are the author's and not necessarily those of the Brookings Institution, its trustees, or other staff members.
The Coalition for Advertising-Supported Information and Entertainment, which sponsored this paper, is a joint effort of the Association of National Advertisers, the American Association of Advertising Agencies, and numerous leading advertisers and advertising agencies across the nation.
The Growth of Communications Media
I define ''communications media" as all of those delivery systems through which information and entertainment are transmitted to and among businesses and consumers. These include traditional print media (newspapers, magazines, journals, etc.), broadcasting, cable television, satellites, and traditional and newer wireless forms of telephone service. Historically, these various media occupied somewhat independent markets in the United States. The print media provided information and advertising aimed at specialized national markets or at decidedly local markets. Radio broadcasting provided entertainment services for which it had little direct competition other than records, but it competed with print media in disseminating local and national news. When television broadcasting emerged, it began to offer entertainment that competed with motion picture theaters, but it too became an important distribution medium for national and local news. Cable television developed as a simple form of television-broadcasting retransmission, butwhen allowed to develop without regulatory controlsrapidly became a medium for distributing a wide range of entertainment and information services. More recently, direct satellite transmissions have begun to compete with cable television and broadcasting in the distribution of national entertainment and information services.
Virtually all of the above media have developed with at least some form of advertiser support. In the case of cable television, it was the FCC's decision to allow cable operators to import distant advertiser-supported signals that provided the impetus for cable systems to expand into large urban markets and to increase their channel capacity. These expansion decisions, in turn, facilitated the development of new basic cable networks that were also supported by advertising.
The telephone industry, on the other hand, has traditionally been supported by direct customer access and usage payments. The only source of advertiser support for traditional telephone carriers has been from the publication of Yellow Pages directories. This reliance upon direct customer support was dictated by the nature of the service: two-way switched voice communications. It would have been difficult and disruptive to have advertisements interspersed with these conversations. However, with the development of computer connections through modems or local-area networks, the delivery of information and entertainment services over telephone circuits became possible. As with other communications services (other than voice telephony), customers may now be offered a choice: pay for these services directly, accept advertisements in them, or accept some combination of the two. For example, electronic Yellow Pages may be advertiser supported, require a subscriber usage charge, or both. Similar options may be available for electronic want ads, home-shopping services, or even electronically delivered financial services.
The role of advertising revenues in the growth of the various media is evident in Table 1. Since 1970, there has been substantial growth in all media that offer advertiser-supported content. Video markets, in particular, have grown substantially, but even the print media have enjoyed an expansion of total revenues from advertising and direct circulation fees. In recent years, multichannel video media, such as cable television and direct broadcast satellite (DBS), have grown more rapidly than traditional radio-television broadcasting. In the next few years, as additional DBS services and even terrestrial multichannel wireless systems are built, video service revenues are likely to continue to grow, supported by both advertising and direct consumer outlays.
For more than a decade, regulation restricted the growth of pay or subscription television services in order to protect off-the-air broadcasters. This, in turn, limited cable television to the retransmission of advertiser-supported off-the-air television broadcasting. This repressive regulation was slowly eliminated by the courts, the FCC, and the Congress in the period from 1977 to 1984, spurring cable to grow very rapidly and to expand its channel capacity.
Even though cable subscription revenues grew very rapidly, however, broadcasting revenues continued to expand, in part because cable retransmissions made marginal UHF stations more viable, a result predicted by academic studies of the medium.2 Thus, cable subscription growth induced further growth of advertiser-supported broadcasting. But even though cable operators would be able to survive and even expand by relying solely on direct subscriber revenues, they also found that they could now tap advertising markets to grow even more rapidly. Direct cable television advertising expenditures began to grow very rapidly in the 1980s, increasing from virtually nothing in 1980 to nearly 20 percent of total cable revenues in 1993.
There was a problem loading page 158.
The telephone industry, supported almost exclusively by subscriber payments for conveying voice and data messages, has grown much more slowly. Telephone companies have not moved aggressively into the provision of content, in part because of regulatory policies that have excluded them from some of the most important of these markets. As a result, their growth has been much slower than that of the print and video media. Indeed, in real (inflation-adjusted) terms, telephone revenues increased at a rate of only 1.6 percent per year from 1980 through 1993. (See Table 2.) During the same period, broadcasting and cable revenues rose at an annual real rate of 6.3 percent per year, and newspaper and magazine revenues rose at a real rate of 1.9 percent per year.
Obviously, telephone companies recognize that their growth has slowed dramatically and have therefore moved aggressively to seek relief from regulation that has kept them from offering new services. In this pursuit, the regional Bell operating companies (RBOCs) have won court reversals of the restriction on their provision of information services that was part of the decree breaking up AT&T.3 In addition, invoking the First Amendment, several RBOCs have recently mounted successful First Amendment challenges to the legislative ban on telephone company provision of cable television service in its own region.4
New Communications Technologies
In the next few years, cable television operators, telephone companies, and satellite operators will be facing off in the competition to build new broadband distribution networks. These new networks, utilizing a mixture of wireless and fiber technologies, may offer a variety of switched and distributive broadband services. At this juncture no one knows which technologies will ultimately succeed, no which services will propel these new technologies.5 This is a risky environment for private investors and policymakers alike given the magnitude of the necessary investments.
Cable television operators now have one-way "tree-and-branch" broadband networks composed of fiber optics and coaxial cables. These networks are typically unable to offer switched services to most subscribers, although such services are possible with major investments in switching equipment and changes in network architecture. To upgrade cable networks simply to provide switched voice (telephone) services would probably cost $300 to $500 per subscriber.6 To provide full interactive switched video might cost as much as $800 per subscriber depending upon the amount of fiber optics already in the cable system's feeder plant.
Telephone companies are now exploring technologies for expanding the bandwidth of their networks so that they can offer video dial-tone or full switched video service. Many companies are pursuing a fiber/coaxial cable architecture that is likely to cost $800 to $1,000 per home passed.7 A full fiber-to-the-curb architecture might be more expensive but provide greater flexibility in terms of bandwidth available for one-way and interactive switched video applications. Simply upgrading the existing copper wire distribution plant with asynchronous digital subscriber loop (ADSL) technology8 is much more expensive per subscriber, but allows the telephone company to target its investment at only those households that wish to subscribe to its video services.
A recent entrant into the video distribution market is high-powered satellite delivery to the home. This direct broadcast satellite (DBS) service has been launched by Hughes Aircraft and Hubbard Broadcasting and
will soon be emulated by two other consortia, including one that now operates with a lower-powered C-band service. Each of these enterprises requires about $750 million to $1 billion in initial investment plus possible start-up losses.9
Other wireless services are attempting to compete with cable systems for one-way, distributive video services, but none has yet been shown to be a viable launching pad for the interactive broadband network of the NII variety. It is possible, however, that terrestrial wireless services can be adapted for this purpose, particularly if the FCC shifts to using market mechanisms for all spectrum allocations. Multichannel multipoint distribution services (MMDS) or cellular cable systems could have sufficient channel capacity, particularly with digital signal compression, to offer competition to wired systems. Finally, it is possible that a system of orbiting satellites, such as that being developed by Motorola for narrowband communications, could also provide switched broadband services.
This brief review of the technical possibilities for building the NII is not intended to be comprehensive. Rather, it is included to show that the number of competing technologies is growing and that the ultimate winner or winners is (are) not known or even likely knowable at this time. Some telephone companies are reexamining their earlier commitments to a fiber/coaxial cable technology.10 At least one large cable television/telephone company merger has failed, and other consortia may be foundering. DBS systems may have arrested some of the momentum for building interactive terrestrial broadband networks because the latter would probably have to rely heavily on traditional one-way video programming at first. However one interprets the recent technological and economic developments in this sector, it is quite clear that building versions of the NII is not becoming less risky.
Even at costs as low as $800 per home, the cost of extending the NII to all residential households is at least $80 billion. To extend it to all business and residential customers would require at least $120 billion.11 This latter estimate is almost exactly equal to the book value of the entire net plant of all regional Bell operating companies and about 80 percent of the book value of the net plant for all telephone companies in the United States.12 Thus, if the NII is to be built by established telephone companies with technology now under development, it would probably require a near doubling of these companies' assets. Since no one is seriously advocating government support that would cover even a small fraction of this added investment, private firms must have the ability to obtain large cash flows from a variety of sources to fund such a massive expansion of our communications infrastructure.
The sobering fact about any such large investment in the NII is that it will likely be overcome by different or improved technology soon after it is built. Perhaps such networks can be continuously adapted to these technical improvements, but it is also possible that a completely different technologyperhaps based on orbiting satellites, for examplecould render an $80 billion to $120 billion terrestrial investment obsolete within a decade. Under these circumstances, investors should be permitted to explore all possible sources of revenues from the marketplace if we expect them to commit such large amounts of capital to so risky an enterprise.
Recent Trends in Communications-Sector Capital Formation
The explosion in new technology might be expected to generate a boom in investment in the various communications sectors. The advances in fiber optics, electronic switching, asynchronous switching, digital compression, and consumer electronics are providing all market participants with the opportunity to deliver a multitude of new services if they invest in new equipment.
The evidence on capital formation in the communications sector only partially reflects this rosy assessment. Telephone companies responded to the new opportunities by investing in new switching systems and fiber-optics distribution systems through the early 1980s. Much of this investment was spurred by competitive entry into long-distance markets and even some urban local-exchange markets and by the necessity for local carriers to provide equal access to these new competitive carriers in their switching systems. Thereafter, however, telephone company investment slowed markedly. (See Figure 1.) In the past 5 years, investment in this sector of the economy has been insufficient to maintain the net value of its capital stock.13
By contrast, the radio-television sector continued to expand its net capital stock through 1993. Fueled by cable deregulation in 1977–79, the net capital stock in this sector nearly doubled between 1979 and 1989a
period in which telephone plant increased by less than 20 percent. (See Figure 2.) This growth reflects the substantial expansion in the number of homes wired by cable in the 1980s as well as the investments in increasing channel capacity during these years.
A lesson to be learned from these divergent trends in capital formation is that regulatory constraints that restrain competition and new entry, whether in the name of promoting "universal service" or "fairness," create a substantial risk of slowing investment and the spread of new technology. Cable television investment accelerated after the severe regulatory restrictions on cable service, motivated by the desire to protect off-the-air broadcasting, were lifted in the late 1970s. The removal of these restrictions caused cable revenues to rise rapidly, fed by both advertising and subscriber fees, while traditional broadcast advertising revenues continued to rise. Telephone investment rose rapidly in the 1970s after entry restrictions on long-distance competitors were lifted. Once restrictions were placed on telephone company offerings of information and cable television services, however, investment in traditional telephony slowed dramatically. Were new investments in the NII to be constrained by the desire to cross-subsidize new universal-service offerings or by a concern that advertiser-supported services
might supplant other services, the prospects for rapid deployment of risky new technology could be severely impaired.
Advertising Versus Subscriber FeesThe Implications for Economic Welfare
Economists have studied the economic welfare implications of using advertising or direct subscriber payments to support radio and television programming. The results of this research suggest that a mix of the two support mechanisms is likely to be the optimal source of revenues.14 The reasoning is fairly straightforward.
Advertisers are likely to be motivated by the desire to maximize their reach among a target population at the lowest possible cost consistent with persuading consumers of the attractiveness of their products. For this reason, they are likely to eschew the support of programming (or other software) that appeals to a narrow set of potential consumers. Thus, advertiser-supported services are likely to be less finely differentiated than are subscriber-supported services under most circumstances. If some consumers are not willing to pay much for highly differentiated services, their economic welfare will be greater with advertiser-supported services. It is for this reason, presumably, that more than one-third of all households now passed by cable television do not currently subscribe despite the fact that cable offers much more choice than the available array of broadcast stations in even the largest markets.
On the other hand, some consumers may desire more choices than those that can or would be supported totally by advertising. Subscriber-supported services can reflect the intensity of consumer preferences, not simply their willingness to use a servicethe criterion that is important to advertisers. It is for this reason that pay-cable and pay-per-view services have proliferated in the current multichannel video environment. Only a few hundred thousand subscribers, each paying a substantial one-time user fee, may be required to support two or three hours of a nationally distributed service or event. Foreign soccer games, obscure older feature films, or live concerts may be offered for a one-time charge, while more popular events may be advertiser supported.
A third possibility for supporting services on the current or future information infrastructure is a mixture of user charges and advertiser support for the same service. Pay-per-view services could still carry advertising billboards or even interspersed advertising messages. For example, videocassettes and motion picture theaters are now supported by both subscriber charges and advertising. In addition, the sam programming could be offered simultaneously without advertising but with a subscriber fee and with advertising and no subscriber fee. This combination of support mechanisms could lead to greater diversity and a larger array of available services.
It is important, therefore, that policies for developing the NII should not directly or inadvertently exclude new service options that might contribute importantly to the cash flows necessary to amortize the large investments required to build the infrastructure. As long as there are no prohibitions on advertiser support or direct subscriber payments in a competitive marketplace, there is little reason for policymakers to attempt to prejudge the choice of support mechanisms for the current information infrastructure or for the NII. It can always be argued that virtually any information or entertainment service has attributes of a public good. One consumer's use of such a service may not reduce the amount of it that is available for other consumers; therefore, no consumer should be excluded from consuming it through a price that is above zero. Advertising could help provide a solution to this problem since it can be employed to generate free broadcasting or print services with low subscriber fees, for example. However, without a combination of advertising and direct subscriber fees, there is no mechanism for consumers to indicate the intensity (or lack of intensity) of their preferences.15
Regulating the NII
It is impossible for anyone to understand fully the potential of the changes that are now occurring in telecommunications technology. Even if the technical choices for network design could be forecast for the next 10 or 20 years, no one could possibly predict how these technologies would be used. The potential changes in the delivery of financial, shopping, entertainment, medical, and government servicesto name a fewthat are likely
to occur are so vast that they cannot possibly be forecast with any precision. But the ideal design of new networks will be highly influenced by both technology and the evolution of and demand for these new services. Therefore, investments in new networks will necessarily take place on a trial-and-error basis.
In this environment, it would be counterproductive for government policymakers to attempt to guide investments or to require carriers to offer services to segments of the population at nonremunerative rates in the name of "universal service." Such demands can only be made on carriers that are protected from competition in other markets because intra-firm subsidies require that the firm earn supra-competitive returns somewhere. If certain (early) NII services are selected as the domain for universal-service requirements, the protection afforded the carrier(s) in other markets may inhibit entry and the introduction of new technology elsewhere. This was precisely the outcome of the FCC's protection of television broadcasters from the development of cable in the 1960s and 1970s and of the federal/state protection of telephone companies from competition in customer-premises equipment, long-distance services, and wireless services.
The calls for preferential access to the NII from a variety of worthy public institutions are understandable, but they should be resisted if they require carriers or other service providers to cross-subsidize them from other service revenues. If there is a legitimate case for subsidizing some of these institutions' access to the NII, it would be far better to provide these subsidies directly from public tax revenues. Otherwise, the built-in cross-subsidies will serve as the rationale for barring new entrants with even newer services in the years to come. If there is one lesson to be learned from past exercises in regulatory protection in this sector, it is that such protection delays new services, such as cable television, cellular telephony, personal communication service, video-on-demand, or direct broadcast satellite services. Free entry to new facilities, new services, and all sources of revenues will provide the fastest route to the NII.
1. For example, Japan is now urgently considering new approaches for allowing competitive entry into its telecommunications markets, including the possible emulation of our approach to breaking up AT&T. The European Community, which featured entrenched Postal, Telegraph, and Telephone authority (PTT) control of its national telephone monopolies, is now moving to liberalize its telecommunications sector by January 1998. Australia has already admitted a second telephone carrier, Optus, and New Zealand has not only privatized its national carrier and opened its market to competition but has also completely deregulated telecommunications. Finally, Canada has recently admitted entry into its long-distance sector and is considering the liberalization of local telephone service and video delivery services.
2. R.E. Park was the first to show that cable television would increase the viability of UHF broadcast stations. See R.E. Park, "Cable Television, UHF Broadcasting, and FCC Regulatory Policy," Journal of Law and Economics, (1972), Vol. 15, pp. 207–31.
3. This restriction was one of the line-of-business restraints on the RBOCs that were built into the antitrust decree that was entered in 1982 to break up AT&T. The information-services restriction was removed in 1991. (See Michael K. Kellogg, John Thorne, and Peter W. Huber, Federal Telecommunications Law. Boston: Little, Brown and Company, 1992, Section 6.4, for a detailed history of this restriction.)
4. The successful challengers have been U S West, Bell Atlantic, Southwestern Bell (now SBC), NYNEX, Southern New England Telephone, GTE, BellSouth, and Ameritech. These cases have been brought in various U.S. District Courts and, in two cases, U.S. Courts of Appeals. The Supreme Court has yet to rule on the issue.
5. An excellent example of this uncertainty was provided by Bell Atlantic in the past month. It has withdrawn its Section 214 applications for the hybrid fiber/coaxial cable and ADSL technologies, announcing that it now wishes to pursue the fiber-to-the-curb technology and various wireless technologies, such as MMDS.
6. See David P. Reed, Putting It All Together: The Cost Structure of Personal Communications Services. FCC, OPP Working Paper #28, November 1992, p. 35, and "Will the Broadband Network Ring Your Phone," Telephony, December 6, 1993, p. 34.
7. This estimate is based on recent filings by Pacific Bell before the Federal Communications Commission (see, for example, Robert G. Harris, "Testimony in Support of Pacific Bell's 214 Application to the Federal Communications Commission," December 14, 1994). Slightly higher estimates are to be found in David P. Reed, Residential Fiber Optic Networks: An Engineering and Economic Analysis. Boston: Artech House, 1992.
8. This technology allows telephone companies to deliver video services over the paired copper wires that are now used to connect subscribers to the companies' plant.
9. See "Digital TV: Advantage, Hughes," Business Week, March 13, 1995, pp. 67–68. Similar estimates may be found in Leland L. Johnson and Deborah R. Castleman, Direct Broadcast Satellites: A Competitive Alternative to Cable Television? RAND, 1991.
10. Bell Atlantic has recently withdrawn its Section 214 applications for hybrid fiber/coaxial cable technology but remains committed to building a broadband network using other technologies, such as fiber to the curb.
11. This estimate is based on a national total of about 150 million access lines. (United States Telephone Association, Statistics of the Local Exchange Carriers, 1993.)
12. FCC, Statistics of Communications Common Carriers, 1993/94 edition, p. 38.
13. It should be noted that part of this slowdown is attributable to the liberalization of customer premises equipment. Since the late 1970s, customers have been able to purchase their own telephone handsets, key telephone systems, PABXs, modems, answering machines, and fax machines. As a result, perhaps 20 to 25 percent of all telephone-related capital equipment is owned by telephone customers. See R.W. Crandall, After the Breakup: U.S. Telecommunications in a More Competitive Era. Washington, D.C.: The Brookings Institution, 1991.
14. For an excellent review of this literature, see B.M. Owen and S.S. Wildman, Video Economics. Cambridge, Mass.: Harvard University Press, 1992, ch. 4.
15. For an excellent compendium of studies of the public-goods problem, see T. Cowen (ed.), The Theory of Market Failure. Fairfax, Va.: George Mason University Press, 1988.