Viewpoint The World Bank GroUp JUly 1997 Note No. 119 Telecommunications Is Dead, Jfnaes Bond( Long Live Networking The effect of the information revolution on the telecom industry Powerful forces are Economic history teaches us that no industry The threat to the giants arises from the huge de- recasting the business is immune to change. Canals gave way to rail- cline in the cost of communicating, the increase world in a fleeter, more competitive formr roads, which in turn bowed to road transport in the power of computing, and the shift to digi- These forces, largely when the cost of motor vehicles fell enough to tal technology (Viewpoint 118). These forces have grouped around make it more cost-effective. Banks are scram- led to industrial convergence as communications information infra- structure and new bling to adjust to a world in which debt can be and information services (such as basic telephone communications raised on bond markets and consumers can service and cable TV) have been delinked from technologies, have obtain many traditional banking services on- their delivery infrastructures (the telecommuni- come to he kniown collectively as the line. And the telecommunications giants, which cations and cable networks). With the delinking information revolution. have reigned supreme for the past fifty years, has come increasing overlap between the two This Note is the second are being besieged in their turn as the infor- main components of the communications indus- in a series of five that looks at this revolution mation revolution overturns the certainties on try: common carrier conduit systems (delivering and the future of which their strength is based. telephony) and content-based information telecommunications. sources and technologies (broadcasting). As a result of these changes, newv competitors are emerging from unexpected directions, and the market domination on which telecommunications companies base their strength is melting away. I TELECOMMUNICAnONS INFASTRUCIURE has three main components: The economics of telecommunications tenminal equipment, such as telephones and fax machines in users' homes and businesses; the local loop, generally a pair of New technology has profoundly altered the copperwiresconnectingtheterminalequipmenttoswitching industry's cost structure, and as a result, the equipment in the local exchange; and long-distance or interna- existing structure of the industry and its pric- ing methods have nowv become incoherent. The tional transmission networks, made up of fiber-optic cables, end ofsnatr nopoly,ote tndoward ne microwave links, and satellites.end of natural monopoly, the trend toward ne", pricing structures, and the increasing competi- * TeEcowAuIAIos sERv cEs have focused on calling services: local tion and globalization in the industry are forc- calls (within the local exchange network) and long-distance or ing radical change. international calls. Increasingly, however, these basic services have been augmented by the transmission of data in binary form The end of scale and natural monopoly and by value added services (such as call waiting and Internet In a conventional wired network, most of the access), which increase functionalityforthe end userand investment goes to establish the local loop, par- generate supplementary income forthe telecommunications ticularly the civil works needed to extend the operator. network to the end user. About two-thirds of the assets on the balance sheet of telecommu- nications operators are holes in the ground"- -J 1Industry and Energy Department * Finance, Private Sector, and Infrastructure Network Telecommunications Is Dead, Long Live Networking FIGURE 1 BURIED COPPER WIRE GIVES WAY TO WIRELESS IN LOCAL LOOP AS TECHNOLOGY DRIVES COST DOWN: AS TECNOLGYRIVSCSTOWNthe trenches needed to lay the cables. Thus, in Annual lifetime cost (U.S. dollars) a traditional network, 70 to 85 percent of the 1,200 Buried copper wire cost of a call, eveni an international one, con- 1,000 sists of the cost of the low-technology link cov- 800 ering only the last couple of miles. 600 Because in the conventional local loop based 400 4f on copper wires the marginal cost of each new 200 subscriber declines no Imatter how manv existing Wireless subscribers there are, the telecommunications 0.1 1 10 100 1,000 sector-or at least its local loop portion-has Subscriber density (lines per square kilometer) been considered a natural monopoly. Economic (log scale) theory suggests that the best way to manage a Source: Coopers & Lybrand data; European Bank for Reconstruction and Development natural monopoly is to create a regulated util- data. ity, granting it a franchise to deliver the service in exchange for certain obligations (such as non- FIGURE 2 WORLDWIDE EXPLOSION OF NEW SERVICES discriminatory treatment of consumers) and regulating the prices it can charge for the end Index (1990 = 1001 product. This explains the nearly universal model 1,500 for the telecommunications sector: a local Internet subscribers(94.2%) monopoly company, often a public enterprise, 1,000 with regulated prices. But wireless, cable TV, and other technologies 500 Cellular (51.5%) are now challenging the conventional local loop based on wireline technology and buried cop- Tlehone lines (5.8%) per (figure 1). In many cases, wireless is already cheaper per new subscriber than wireline. And 1990 1991 1992 1993 1994 the much flatter cost curves of wireless show that Note: Figures in parentheses are annual growth rates overthe period shown. size no longer brings any real cost advantage. It Source: World Bank and International Telcommunication Union data. is possible to have several competing providers of local service without raising the network's overall costs much. The implications are consid- FIGURE 3 EXPLOSION OF NEW TECHNOLOGIES erable: the best way to deliver service to custom- 1997 Voice pager ers is no longer through a utility but through 1996 | Networked PC 1995 Fixed wireless competing providers of local telecommunications Digital TV services. The existing telecommunications sector Asynchronous Transfer 1980 Personal computer model in most of the world is simply wrong. Mode 1918 Compact disc 193 Gyblpsitioin 1971 Chip Furthermore, the telecommunications reforms 1991 Internet explosion 1970 Fiber-optic cable sweeping the world should focus more on the 19I ISDN Laser 1986 Local area networks 1969 Computer network structure of the sector-providing as much 1985 Cellular telephone 1960 Satellite for as on 1947 X}TT Transistor potential competion possible-than 1945 Computer the transfer of the monopoly telecommunications 15 - aTelevision company from the public to the private sector. 1906 Radio In many cases, however. the reverse is true. 1876 -Telephone ___ _______________- ____________ Telegraph Telegraph 101520The move to bandwidth-based pricing 1850 1900 1950 2000 Source: World Bank compilation. Source.-World Bank compilation. Almost universallyv how much you pay for a telephone call depends on how long you talk BOX 2 THE IMPENDING COLLAPSE OF THE INTERNATIONAL ACCOUNTING HATE SYSTEM and how far away your correspondent is. If you are in Paris, it is much more expensive to call International tariffs are based on the accounting rate system, which New York than to call Toulouse, and the price was developed as part of a regulatory tradition holding that interna- you pay (beyond the monthly rental fee for the tional telecommunications services are supplied through a bilateral line) is proportional to the time you talk. But correspondent relationship between national monopoly carriers. An that is not how costs are built up for the opera- tortis ino theosecosto ore example, tarf for itera- accounting rate is the price the two national carriers (or their tors in the sector. For example, tariffs for inter- gvrmtt)ngtaefrhnln n iueo nenfoa national calls, based on the current but outdated governments) negotiate for handling one minute of international accounting rate system, nearly always far exceed telephone service. Revenues are shared between the two carriers. the cost of providing the service (box 2). The accounting rate system was originally intended to allow each carrier to recover its costs for handling an international call. A second problem relates to new services. The The main problem with the accounting rate system is that for new communications services that customers in- creasiy dmmucaLond sendvaryins thamcuntoms oin- nearly every country the cost of transmitting a call has fallen formation per second down the transmission dramatically over the past twenty years, but the fall in price has line. Paging, for example, requires narrow band- lagged this decline. As a result, the rate greatly exceeds the cost of width (a small amount of information per sec- providing the service, so accounting rates, which still assume that ond), while new multimedia services (such as the sector is a monopoly, generate huge economic rents for tele- teleconferencing) require a huge amount of phone companies handling international calls. In some cases, bandwidth because transmittino video sends much more informaiondontheinethinternational calls account for the entire profit of the sector and even much more information down the line than does transmitting sound alone. But most telecommu- generate foreign exchange for the government nications operators do not offer choice in band- This system is showing signs of imminent collapse. New possibili- width: customers get a standard telephone line, ties for competition in international services make the sector a fertile accommodating 64 kilobits per second (kbps) area for arbitrage-and so we are seeing significant activity in call- in Europe and 56 kbps in the United States. back services, calling cards, Internet telephony, and the like. These To an increasing degree, the costs borne by' new sources of competition undercut the hugely inflated accounting telecommunications operators are made up of rates and eat into the income of telecommunications operators- three elements: a fixed monthly amount, which especially those offering the lowest prices. As a result the U.S. corresponds to the capital costs of the local regulator, the Federal Communications Commission, is seeking to loop; a one-time cost for each connection, cor- replace international accounting rates with a new benchmark responding to the cost of switching that call; system based more closely on actual costs. and a transmission cost, which is proportional to the bandwidth. But actual customer charges are quite different. And because it is increas- ingly possible to compete for customers, new, will probably be able to buy the bandwidth agile operators are emerging that take advan- capacity they need for a given connection, tage of the possibilities for arbitrage between which, because of declining transmission costs, tariffs aril actual costs. Much of tlle new activ- will cost nIo mIore thaIn a few cents per hour ity is in international service. But new players even for international calls (in addition to the are also emerging in other areas, such as Inter- monthly rental for local access). net service providers, which enable customers to place long-distance calls for the price of the Intermodal competition, globalization, local loop connection (figure 2). and the WTO negotiations It is safe to predict that competition and mar- Policymakers increasingly accept that competi- ket forces will drive tariffs closer to long-run tion in the local loop is both possible and desir- incremental costs, both in level and in struc- able. But competition is also bringing changes ture. And within a couple of years, consumers that policymakers are much less knowledgeable z1 ~~~Telecommunications Is Dead, Long Live Networking about-in the explosion of new technologies, nications markets to foreign investors and op- products, and services competing with one an- erators has important benefits both locally and other to deliver connectivity to the end user (fig- globally. The WT O has placed telecommuni- ure 3). New technologies increase arbitrage cations at the top of its agenda for multilateral possibilities for new operators, and they compli- trade liberalization (Viewpoint 120). As mar- cate the work of regulators. But above all, they kets open to foreign participation, and as tech- challenge existing operators, which are often slow nology creates new markets, we can expect to in responding to new customer demands. see an entirely new cast of players investing in markets that are new not only technologically Also changing is the geographic service area but also geographically. in which end users are interested-increasingly not only national but international. International Conclusion traffic is growing by 12.4 percent a year, com- pared with 5.9 percent for domestic calls. Thus, The old monopoly telecommunications sector as competition in domestic markets becomes is fast disappearing. N\ational telecommunica- the norm, consumers will find operators that tions markets are fragmenting into a multiplicity c an offer packaged services on an international of niche markets at the same time that national scale increasingly attractive. The growing de- trade barriers are falling. Many new operators mand for international services helps explain are emerging, each targeting the segment that the trend towTvard global alliances among tele- best corresponds to its comparative advantage. Viewpointeisdan tope communications operators. But this trend may In this new networked bit industry, offering a eocoarage dissemina- stem more from the desire of former monopoly huge range of competing technologies and ser- tion of and debate on players to recreate at the international level the vices, the future for the incumbent telecom- ideas, innovations, and best practices for oligarchies to which they are accustomed in munications companies looks increasingly expanding the private their domestic markets than from the underly- bleak. Over the next decade, as the market sector. The views ing market forces in the industrv, shifts from under their feet and as new, more published are those rf the authors and should nimble actors emerge, we can expect the domi- not be attributed to the In an alternative future, transmission capacity nance of the telecommunications operators World Bank or any of its and bandwvidth might become tradable com- over their traditional markets to erode spec- affiliated organizations. Nor do sany of the cor- modities, with a spot market on which capac- tacularly. Some countries are likely to see their closions represent ity is bought and sold in half-hour slices and major telecommunications operators fail, and official policy of the megabit-per-second tranches. The spot market new players will appear that quickly become WorltivDBakoofiretos could be associated with a futures market on household names around the world. Or the countries they which contracts would allow operators to hedge represent. future positions. It is difficult to imagine today's James Bondt (jbond@Xworldbanle.org), Division To order additional mastodon operators, which now cover every- Chief, Telecomm^unications and Infonnatics copies please call 282- thing from bulk transmission to value added 458-1111 or contact services, in such a world. Instead, the world Suzanne Swish, editor, would segment into wholesalers, which w ould The World Book, invest in and sell capacity; retailers, which 1818 H Street, NW, W ould be in contact with the final consumer Washington, D.C. 28433, and traders and brokers, which would inter- ssmithj©Zworldba nk.org. mediate supply and demand for capacity. The series is also available on-line (www.worldba nk. But globalization goes beyond relecommuni- org/html/fpd/notes/ cations operators. As Lhe World Trade Orgarii- noste list. html)I. zation (WTO) focuses world attention on @ Printed on recycled liberalizing trade in services, policymakers are paper. beginning to realize that opening telecommu-