Eight
Race
Making Connections

With ESPRIT on track and out of the station, the Commission began to fire up the engine for its other telematics vehicle, a program in telecommunications. Its handling of ESPRIT had won for the ITTF a degree of credibility as an administrator of large-scale, industrially important, high-technology research. The time was right to capitalize on its favorable image and on the nascent optimism regarding European telematics collaboration that it had inspired. The telecommunications program, eventually called RACE, would follow the same pattern that had made ESPRIT a success: Start with an industrial coalition, then win the approval of governments. The recipe was a good one, though more complicated in the telecommunications sector.

As in semiconductors and computers, Europe's telecommunications-equipment industry was the domain of national champions. In fact, the same companies were sometimes prominent in both IT and telecoms, companies like Siemens, Philips, and GEC. A complication for the telecommunications sector was that it had long been the administrative demesne of the public telephone administrations in each country. The PTTs, with their historic monopolies on all aspects of telecommunications in Europe, added a new dimension to the Commission's task in RACE. In the end the PTTs approved of RACE after reshaping it somewhat to their liking. The

key analytic factors in explaining the movement from crisis to collaboration are the same as those that figured in ESPRIT: an entrepreneurial IO, a powerful transnational industrial coalition, and national governments in the midst of profound policy adaptation. In the present chapter I first describe the crisis in telecommunications facing European governments. I then detail the political processes that led to RACE.

Turmoil in Telecoms

The telecommunications realm has been in upheaval since the late 1970s. Much of the tumult stems from the microelectronics revolution. New electronics technologies have led to increasingly capable and efficient communications equipment, which makes possible myriad new services, which in turn strain the capacities of traditional institutional arrangements—namely, the PTTs. All these developments have been, and continue to be, exhaustively analyzed and commented on. I will therefore summarize as succinctly as possible the key technological and regulatory changes that have placed pressure on Europe's traditional telecoms structures.^[1]

A Changing World

Telecommunications involves the creation of electronic links between two locations for the purpose of sending and receiving information. In early voice telephony, the most basic of telecommunications services, an operator sitting in front of a switchboard established the connections by plugging a jack into a socket. After World War II telephone connections were created by electrome-chanical switches, in which an electrical impulse triggered the movement of a metal reed. Progress in microelectronics in the 1970s

[1] The revolution in telecommunications has been so dramatic that virtually innumerable books and studies on the subject have emerged. Most of what follows is general background knowledge, though I use primarily the following sources: Borrus et al., Telecommunications Development; Gilbert-François Caty and Herbert Ungerer, "Les télécommunications: nouvelle frontière de l'Europe"; Hart, "The Politics of Global Competition"; Jill Hills, Deregulating Telecoms: Competition and Control in the United States, Japan and Britain; Nguyen, "Telecommunications"; and OECD, Telecommunications .

made possible the development of fully electronic switches. In these switches the connections are made within the ICs themselves, involving no moving parts. Software programs control the switching. Thus, at one level today's telecommunications exchanges resemble computers: They are made of thousands of ICs and are controlled by preprogrammed series of instructions. Such "digital" switches can operate on digital or analog networks.^[2]

New transmission modes have also revolutionized communications. Microwave relays permit transmission without the need for laying cables. Satellites can perform a range of telecommunications functions by making point-to-point connections, linking several locations, or even broadcasting (point-to-multipoint communications). The emerging generation of transmission facilities capitalizes on fiber optics. Glass filaments carry information in the form of pulses of light. The advantages of fiber optics are enormous: greater capacity (a single filament can carry three times more voice conversations than can a current coaxial cable), immunity to electromagnetic noise and interference, and lower rates of signal attenuation (meaning fewer repeaters). Plus, the fibers are made of one of the earth's most abundant substances, silica.

The new switching and transmission techniques have made possible a proliferation of services beyond basic voice conversation. As computers spread throughout business and industry, the need to transmit vast quantities of data ballooned. New data-processing companies, like Electronic Data Systems, depended on telecommunications links to receive inputs and to return data after processing them in their large mainframes. Numerous industries now require constant, reliable, high-volume data communications—for example, banks clearing their accounts or airlines processing their worldwide reservations and ticketing. Other new services include

[2] Electronic switching is usually called digital because all the commands for running the exchange exist in the form of binary electronic bits—the same basis on which computers operate. A digital switch can, however, route traffic through an analog telephone network. In such "space-division switching," the transmission facilities do not carry digitized information but rather analog signals, which vary continuously over a range and mirror the modulations in the original message (for example, a voice). The digital switch simply completes the circuits through which the analog signals flow. Time-division switching involves a fully digital network: The original signals are converted into packets of digital information, which are fed into the network and routed through the switch unchanged. They are then reconverted if necessary into an analog form (voice, music) at the receiving end.

paging, mobile phones, videotex,^[3] teleconferencing,^[4] and electronic mail.

Up to now the different kinds of services have required physically distinct networks. For instance, telex, a text-transmission service, uses lines distinct from the telephone network. Presently, voice telephony and low-speed data transmission flow through the basic network. But high-volume data transmission and teleconferencing require circuits with greater capacity. Thus, users with extensive need for rapid data transmission frequently (in the United States, at least) use separate, fully digital networks. In the visual domain cable television employs a completely distinct network.

In the near future that will change. The next step in network evolution will probably be something that is now loosely called ISDN: Integrated Services Digital Network. ISDN will be fully digital (switching and transmission) and will integrate voice, data, and text services. Further in the future is broadband communications, which will likely be based on optical fiber, microwave, and satellite transmission. The increased bandwidth of broadband systems is like additional lanes in a freeway: It can carry more traffic at once. Broadband systems will be able to carry more bits of digitized information; whereas ISDN will carry two Mbit/s (two million bits per second), broadband telecoms should be capable of at least 140 Mbit/s, with some systems currently being designed for 600–1,440 Mbits/s. Broadband systems will permit a single network to carry everything from basic voice conversation to high-speed data to high-quality moving pictures (HDTV).

As technology made possible ever more advanced uses of telecommunications, the demand for new equipment and services took off, especially among major business and industrial users. As Bar and Borrus put it, "In ways never before possible, companies are able consciously to design and build telecommunications networks that decisively enhance their competitive position."^[5] The burgeoning

[3] Videotex is the name for a range of information services aimed at the mass (residential) market. The services are sometimes divided into teletext and videotext. Teletext is one-way transmission of text, such as news headlines or stock-market quotations. Videotext combines text and graphics and is interactive—that is, the user can interrogate the information service and give instructions. For example, with a personal computer and a modem a home user can now go on electronic shopping sprees, ordering from a videotext "store." Or the home user can peruse airline schedules and purchase a ticket herself. Videotext systems are frequently a gateway to dozens of different services. The best-known American videotext systems are probably The Source, CompuServe, and Prodigy.

[4] Teleconferencing allows meetings among groups at geographically distant locations via linkups that transmit both sound and video pictures of the participants.

[5] Bar and Borrus, From Public Access, 1.

demand among businesses for new telecoms services has placed strains on traditional institutional and regulatory arrangements. The pressures erupted in the 1980s into a continuous debate, in public as well as trade forums, over deregulation and liberalization. Policy adaptation was universal. In Chapter 6 I described how aggressive regulatory change in the United States and Japan exerted powerful pressures on Europe's traditional telecoms structures. Freed from past restrictions, IBM and AT&T entered new sectors (telecoms and computers, respectively) and made Europe a prime target for expansion. Japan's well-organized efforts to prepare for broadband communications threatened to place Japanese firms in the lead in the future for equipment and services.

In short both technological and regulatory changes were sweeping across the advanced countries by the early 1980s. Europe's PTTs could not find refuge in their entrenched monopolies. The national governments were (and are, and will be) adapting their telecommunications policies and regulatory structures. The Commission's proposals for European cooperation in managing the changes thus arrived at an extremely opportune moment. Everybody was preparing for an overhaul of telecommunications, but the precise lines of future networks (ISDN, broadband) and institutions could not yet be discerned. The Commission proposals therefore entailed not an upsetting of otherwise stable arrangements but a coordinated, regionally planned management of the changes underway. Of course, that did not mean there would be no disagreements over the direction and rate of adjustment. But national leaders in an adaptive mode were more receptive to new ideas than they otherwise would have been.

Adaptation in Europe

In telecommunications Europe in the early 1980s was not a community but a collection of fiefdoms. Each national telecoms fiefdom was ruled by the administration of posts, telegraph, and telephone. The acronym PTT reveals the origins and nature of the administrative structure: Telegraphs had been added onto the postal services, and telephones were an extension of telegraphs. Telecommunications was seen as a natural monopoly and a public utility. Because telecommunications was considered a natural monopoly, it was believed that only a single supplier of networks and services

could achieve economies of scale. Thus, European PTTs were public monopolies of telecoms networks, terminals, and services. Telecommunications constituted a public utility because, as with roads and electricity, there were social gains from providing universal service. Opening segments of the telecommunications system to competition cuts against the public-utility ideology that has dominated telecoms policies in Europe.

Naturally, the national telecoms administrations resist any such encroachments on their domain, the kind of defense of bureaucratic turf one would expect of any organization. In addition the postal services in most European countries are heavily unionized, and telecommunications revenues have long underwritten postal losses. Thus, the postal unions resist anything that might cause layoffs or cut budgets.

Furthermore, it is by no means clear that deregulation American style is socially or economically optimal in the long run. Deregulation has led to a plethora of different public and private networks in the United States, providing advanced services to that part of the business community that can afford them. At some point the fragmentation may prove a handicap. With much of the telecoms realm in the United States privately owned by large users, innovation may be privatized. The existence of diverse networks may spread out revenues so much that no one can make the huge investments needed for network modernization (toward broadband, for instance).^[6] By contrast, a single, universal network—guaranteed by a state administration—could carry the same services for business and make them compatible with each other but also provide access to advanced services for residential customers.

No one can predict unequivocally which will be the best route to liberalization and modernization—wide-open laissez faire or PTT-led evolution. Because of Europe's administrative heritage, the PTTs will have a major say in planning the transition to next-generation telecommunications. The reforms have begun, and in each country they follow a distinct path. Table 8.1 provides a chronology of the major events in European telecommunications policy.

The United Kingdom

Of the European countries the United Kingdom has proceeded farthest along the path of liberalization. The government commissioned in 1976 a review of the operations

[6] These concerns are explored in ibid., and in Borrus et al., Telecommunications Development , 12–15.

of the BPO, a public corporation that at the time encompassed both postal and telecommunications services. The Carter Report, which resulted (1977), recommended that the postal and telecoms activities be split into two different public corporations, and that the telecommunications network be modernized by rapid introduction of the System X electronic exchange. Thatcher's Conservative government acted quickly on the Carter recommendations, creating a nationalized company in 1981, British Telecom (BT). In December 1984 the government sold 50.2 percent of BT stock in a public offering. The new government also initiated in 1981 a study on the telecoms monopoly of BT. Following the Beesley Report, which resulted

from the study, the government began to liberalize the markets for telecommunications networks, terminals, and services.

The first step was the authorization of a new carrier (network operator), Mercury, in 1981. The new company, now wholly owned by Cable and Wireless, laid fiber-optic cables along British Rail tracks, linking major cities in a fully digital network. Mercury aimed initially at the business market (especially in downtown London) and trunk (intercity) and international communications. Though Mercury remained small compared with BT (total revenues for 1987–88 of about £380 million versus £9,880 million for BT), the threat of competition forced BT to rationalize and formulate new strategies. Competition on network provision also remained limited because the government could not authorize any new carriers before the end of 1990.^[7]

Liberalization touched other important areas of British communications. A new Office of Telecommunications (Oftel) regulates the industry (somewhat as the FCC does in the United States), ensuring competition and fair rates. The market for terminal equipment has been open since 1984, meaning that users can purchase telephones, modems, PABXs, and other devices from competing suppliers. Previously BT had a complete monopoly on customer-premises equipment. Even in switching equipment BT has introduced competition in its purchasing practices. BT has made clear that it will no longer buy solely from the traditional U.K. suppliers, Plessey and GEC (which merged their telecoms interests in April 1988) and STC. In fact BT has already placed some orders with Thorn-Ericsson and an AT&T-Philips joint venture, Pye TMC.

To modernize, BT has announced plans to merge its voice, data, and telex networks into an ISDN. The goal is to have 80 percent of customers connected to ISDN by 1992.^[8] An experimental ISDN, called Integrated Digital Access (IDA), began operating in the fall of 1984. In preparation for ISDN, interexchange lines were fully digitized by 1989. One potential problem is that Britain has chosen an ISDN model different from that favored by other European countries and the CCITT.^[9] The British expect to upgrade their system

[7] Bar, "Telecommunications in the United Kingdom," Appendix, 76–79.

[8] "Grande-Bretagne: des 'Points d'acces,'" 49.

[9] The CCITT, a body of the International Telecommunications Union, formulates recommendations on international telecommunications standards. The British system comprises one channel for voice and data at sixty-four Kbit/s plus two channels for simultaneous data and signaling at eight Kbit/s each. The CCITT proposed standard includes two voice and data channels of sixty-four Kbit/s each and a single sixteen Kbit/s signaling channel.

to fit the CCITT standards, while the current standard allows early use.

British deregulation has placed pressures on continental telecoms authorities. For instance, international telephone calls are significantly cheaper from London than from most other European capitals.^[10] As a consequence, some firms are relocating their communications centers to London. The DGT in France had to adjust its tariff schedule, raising local rates to reduce international charges, in order to keep customers from routing their transatlantic calls through the United Kingdom.^[11]

France

French telecommunications has followed the typical French high-technology policy: state-engineered corporate mergers and ambitious plans. The DGT was until 1987 the French agency charged with responsibility for telecommunications policy; it falls under the Ministry of Posts, Telegraph, and Telephone but generates its own revenues and has a budget independent of the national budget. The DGT, through its research arm, the Centre National d'Etudes des Télécommunications, pioneered research on fully electronic exchanges and in 1972 began installing CGE's E-10 switch, the world's first fully digital exchange.

The DGT also had a hand in industrial policy for the telecoms sector until 1987. The DGT folded ITT and Ericsson subsidiaries into a telecommunications division for Thomson in 1975. The Socialist government nationalized Thomson and CGE (the other main telecoms equipment supplier) in 1981, and two years later merged the telecommunications divisions of both companies into Alcatel, a CGE subsidiary. The DGT opposed that move because it ruined the policy of maintaining competition among suppliers. The DGT also protested when the government diverted a major share of its revenues to fund the filière électronique and to aid the national budget. All of a sudden the DGT lost its surpluses and had to borrow in order to make necessary investments. Alcatel created the world's second largest equipment maker by purchasing ITT's European telecoms subsidiaries in a deal finalized in January 1987. Later, the

[10] Sabine Delanglade and Eric Rohde, "PTT: Déréglementation accélérée en Europe," La Tribune , 26 November 1985, p. 6.

[11] John Wilke, "Can Europe Untangle Its Telecommunications Mess?" 47.

French government sold a small equipment firm, CGCT, to a group including Ericsson (Sweden) and Matra (France). Finally, under the government of Jacques Chirac, responsibility for the telecoms industry was transferred out of the DGT to the Ministry of Industry.^[12]

From 1975 to 1980 the DGT presided over the aggressive expansion of the telephone system to bring the penetration rate up to the level of other advanced countries. It sponsored in 1978 the Plan télématique , which aimed at the rapid introduction of advanced telematics (combining computers and communications) services. The Plan included experiments with a public videotex system, an electronic directory, broadband optical-fiber facilities, and plans for a new telecoms satellite. Transpac, a packet-switched network for data transmission, came on line in 1979.

The French have been among the most ambitious in Europe in planning for advanced services and future networks. In 1987 their level of digitization of the network (55 percent of switches, 70 percent of transmission) was the highest in Europe.^[13] France has had a pilot ISDN project since 1987 at Renan, and by 1988 the telecoms administration was preparing the transition from its separate specialized networks (the four "Trans" systems) to a single ISDN network.^[14] An experimental broadband system has been in operation since 1983 in Biarritz with some 1,500 subscribers.

One of France's great successes has been the Teletel/Minitel system, a videotex service aimed at the mass market. The system offers the DGT's Electronic Directory Service (an on-line telephone directory for the whole country) as well as the Kiosque. Private companies offer information services through Kiosque; by April 1986 1,900 data services were available to professional or residential customers. Use of the Minitel by the public has skyrocketed, largely because of the policy of the DGT to "lend" Minitel terminals free of charge. The terminals were introduced on a trial basis in 1981, and by the end of 1985 1.3 million were in use.^[15] Through December 1987, 3.37 million terminals had been installed.^[16]

[12] Thierry Vedel, "La 'Dérèglementation' des Télécommunications en France," no page numbers.

[13] "France: Le Pays le Plus Numérisé," 49.

[14] See Marie-Laure Théodule, "La gamme Trans en pleine mutation vers le RNIS."

[15] Jeffrey A. Hart, "The Teletel/Minitel System in France," 21–25.

[16] Paul Betts, "A Frenzy of Alliances," Financial Times , 11 May 1988, European Telecommunications Survey, p. 10.

Further changes liberalized several other aspects of French telecommunications. In September 1986 the Chirac government created the Commission Nationale pour les Communications et les Libertés (CNCL). The CNCL was supposed to regulate telecommunications and broadcasting much as the FCC does in the United States; formerly the DGT both ran and regulated the system. Also, the CNCL could authorize private networks for internal use. In September 1987 the government published a new law opening up the provision of VANs to competition. The law did not permit competition in basic voice telephony, which would remain the province of the DGT. Consortia began to form to offer VANs, including one joining IBM, Crédit Agricole, and Paribas. The next step was to open competition in the provision of mobile telephone systems; two groups were authorized. The Chirac government also renamed the DGT France Telecom and spoke of plans to privatize it. But the Mitterrand electoral victory in the spring of 1988 and the opposition of the Socialists and the unions brought an end to these plans. The second Socialist government replaced the CNCL with a Conseil Supérieur de l'Audiovisuel (CSA). The CSA has the responsibility of authorizing large private networks and new VANs. A decree of May 1989 created the Direction de la Réglementation, which functions much like the American FCC; its charter was fully worked out in 1990.^[17]

France, during the period of reorientation after the failure of the early Mitterrand project (see Chapter 7), became a vigorous promoter of European collaboration. The government announced its intention to "work for the creation of a European telecommunications space." In a lengthy article published in Le Monde the minister in charge of the PTT, Louis Mexandeau, called for increased European collaboration and expressed support for the Commission's telecoms initiatives. His statement stressed reciprocal market opening and common European standards.^[18] The shift in France from national-autonomy strategies to a European outlook helped ease the way for the Commission's proposals leading to RACE.

Germany

The Bundespost in the Federal Republic of Germany was for a long time Europe's most stubborn defender of traditional

[17] Benjamin Coriat, "Régime réglementaire, structure de marché et competitivité d'entreprise."

[18] Louis Mexandeau, "Pour une politique européenne des télécommunications," Le Monde , 3 April 1984, p. 35.

PTT roles and monopolies. To be sure, the basic communications law of 1928 and the Fundamental Law of 1949 granted the Bundespost the exclusive responsibility for building and running telecommunications networks. The Bundespost has interpreted this responsibility to include the networks, services, and even terminal equipment.

As late as the mid-1980s private networks were permitted only for in-house use and could not be connected to the public network. The Bundespost retained its monopoly on the sale of the first telephone handset to every customer and until 1988 was the sole supplier of modems. The Bundespost even forbade the sale of computers with internal modems. Any company that wished to sell terminal equipment directly to users (rather than to the Bundespost) had to have each product approved by the Bundespost. For its network equipment purchases the Bundespost relied on a small circle of German firms headed by Siemens, then SEL (the ITT subsidiary now owned by Alcatel), along with Nixdorf and IBM Germany for data communications and VAN equipment.

The Bundespost developed specialized networks to try to meet the growing demand for enhanced services. The Integrated Digital Network includes data transmission, packet-switched data, telex, and teletex. In 1985 the Bundespost was pushing for broad expansion of Bildschirmtext, a videotex system, apparently with only moderate success. In particularly close collaboration with Siemens the Bundespost developed plans for ISDN. Because of a major error in exchange development strategy, Germany lagged well behind in digitization of the public network, the first digital switches being installed only in 1984. The public ISDN, integrating voice and data services, had nevertheless been installed in thirty-nine cities by 1990 and is to be completed nationwide by 1993. The next step will be to integrate narrowband and wideband ISDN (adding videophone and videoconferencing to the public network). Eventually, a universal broadband system based on fiber optics will be in place, adding radio and television to the public network. Trials were underway in 1985 for wideband ISDN in seven major cities under the Bigfon project; the next step was to link the local networks via fiber-optic trunk lines in the Bigfern project.^[19]

[19] Patrick Cogez, "Telecommunications in West Germany," 54–56; "ISDN Makes Strides in West Germany," 22.

Pressures for liberalization from German business and pressures for increased openness to trade from the United States began in the mid-1980s.^[20] The Kohl government appointed the high-level Witte Commission to study telecommunications reform, and it began its work in early 1985. The report finally emerged in September 1987. The Witte Commission majority recommended a set of reforms that fell short of full deregulation or privatization. In fact four members of the Commission issued a separate opinion, arguing that the proposals did not go far enough and that "only replacement of the monopoly with competition at all levels can lead to a market capable of withstanding the future."^[21] The most important recommendations were these:

Moderate as they were, the proposals drew intense criticism. Only the Free Democrats strongly supported liberalization. Some Christian Democrat Länder governments objected to the reforms, even though Chancellor Kohl committed himself to seeing them through. The Social Democrats and Greens opposed the changes. In this position they supported the postal workers' union, which represents

[20] Hart, "The Politics of Global Competition," 187; Peter Bruce, "Bickering Bonn Tackles Bundespost Monopoly with Reluctance," Financial Times , 2 June 1987, p. 2.

[21] "From Bundespost to Telekom," 407.

[22] "RFA: Les recommendations (en substance) de la commission Witte," 91.

463,000 of the Bundespost's 500,000 employees (making the Bundespost the largest single employer in Germany). The union feared "tens of thousands" of layoffs, rising tariffs, and diminished quality of services. Even the Christian Democrat coalition partner from Bavaria, the Christian Social Union, was reluctant.^[23] Still, the changes were approved by the cabinet in May 1988, after court challenges by the unions failed. The Witte reforms became law in July 1988.

Elsewhere in Europe

Telecommunications reform was not limited to the three major countries of Europe. Italy's fragmented system felt the first breaths of rationalization. A government agency, Azienda di Stato per i Servizi Telefonici (ASST, under the Ministry of Posts and Telecommunications), runs the trunk lines. SIP, part of the state-owned STET conglomerate, manages most of the local networks and linkups to customers. A third company, Italcable (also part of STET), handles international connections. The Italian National Plan for Telecommunications (for 1985–94) foresaw open competition for VAN provision and for the supply of customer premises equipment. In 1987 the leadership of the Istituto per la Ricostruzione Industriale (IRI) group (which owns STET) proposed reorganizing STET into an operating company resembling BT. The new STET would include SIP, Italtel, Italcable, the satellite division, and an alliance with a foreign equipment manufacturer.^[24] AT&T formed an alliance with Italtel, but the rest of the plan has not been realized.

The Netherlands, with the most tightly regulated telecommunications system in the EC, also moved toward competition. The Steenbergen Report of 1985 contained a number of deregulation proposals that found their way into a bill passed by parliament in 1987. Under the new law, which took effect on 1 January 1989, the postal services of the PTT and the telecommunications services were split into separate subsidiaries of a new state-owned enterprise, NV PTT Nederland. PTT Telecommunications retains its monopoly on network provision but must compete with private companies

[23] Peter Bruce, "Bickering Bonn Tackles Bundespost Monopoly with Reluctance," Financial Times , 2 June 1987, p. 2; and David Goodhart, "Reforms Firmly on Track," Financial Times , 11 May 1988, European Telecommunications Survey, p. 10.

[24] James Buxton, "Good Progress after a Swift Change of Direction in Italy," Financial Times , 24 October 1983, Survey, p. 12; Alan Friedman, "A Crucial Year," Financial Times , 11 May 1988, Survey, p. 12.

in an open market for terminal equipment and VANs. The Dutch are also planning to convert the network to ISDN after 1995 and to broadband by the turn of the century.^[25]

The report of a Wise Men Commission in Belgium in 1987 recommended changes similar to those planned in the Netherlands. The Wise Men recommended that the RTT (Régie des Téléphones et Télégraphes) be converted into an independent state-owned enterprise. Under legislation proposed in 1988, the RTT will retain a monopoly over the network and over "essential" VANs (though these remained to be defined). The market for terminal equipment will be opened to competition. Belgium also began an exploratory technical study for broadband networks.^[26]

From these snapshots of situations in various European countries, it is clear that telecommunications in Europe was undergoing profound alterations. The changes were both technological, as networks moved toward ISDN and broadband, and institutional, as the PTTs found that they would have to share some of their traditional markets with private competitors.

The Commission and Telecommunications

Into this unsettled atmosphere the Commission launched its trial balloons for European cooperation in telecommunications development. If national telecoms policy-makers had not been in an adaptive mode, Commission efforts certainly would not have achieved anything. As in ESPRIT major European corporations, including makers of telecoms equipment and large users of advanced telecoms services, played a crucial role in winning approval for collaborative initiatives. The Commission/industry alliance proved once again to be an effective starting point. In the story that follows the focus will be on collaborative R&D and the RACE program. But because the Commission advanced broad strategies for the entire telecoms sector (including standards and market opening), I will analyze those other dimensions of Commission strategy so as to place the R&D component in its context.

[25] Catherine Le Bailly, "Fin du monopole aux Pays-Bas," La Tribune , 26 November 1985, p. 6; Laura Raun, "Signs of Vulnerability," Financial Times , 11 May 1988, Survey, p. 10; Terence Holsgrove and Marion Howard-Healy, European Telecommunications at the Crossroads , 7.

[26] William Dawkins, "The Stakes Are High," Financial Times , 11 May 1988, Survey, p. 11.

Evolution of Commission Telecoms Strategy

As described in the previous chapter, Commissioner for Industry Davignon brought to Brussels a keen appreciation of the challenges facing Europe's telematics industries and the need for collaboration. Already in 1979 Davignon was advocating the need for a European industrial strategy to meet U.S. and Japanese competition in telematics. For Davignon telematics meant the broad sector comprising components, computers, and telecommunications. IT (microelectronics and computers) received attention first, in the ESPRIT program. There were reasons for that ordering of priorities. For one, the political problems in telecoms would be much trickier than in IT, involving as they did the PTTs. An initial failure there would put a damper on everything. A second reason for approaching telecoms slowly was that ESPRIT would cover many of the basic microelectronics technologies needed in telecoms systems. Thus, the telecoms program could be a downstream, or applications, program, capitalizing on ESPRIT results.^[27]

The first step, as mentioned in Chapter 7, came in September 1979. Davignon proposed that industry, governments, and the Commission jointly work out an EC strategy for telematics. Regarding telecoms, he suggested that work focus on common standards, on an integrated digital network linking European institutions that would be a pilot for a future universal European digital network, and on encouraging the growth of European data services.^[28]

The following month Davignon sat with the general directors of the telecommunications administrations of the member states as they drafted a set of recommendations on EC telecommunications. The Commission approved the recommendations, which were ratified by the PTT general directors in December. The key points included:

[27] The CEC makes this argument in its ESPRIT I proposal: Proposal for a Council Decision Adopting the First ESPRIT , 11, n. 8.

[28] Agence Europe , 24 March 1979, p. 10; 29 September 1979, p. 12.

[29] CEPT is a body of twenty-six European PTTs. It has traditionally coordinated work on network interfaces and tariff arrangements.

Of course, nothing the PTTs agreed to in CEPT was binding.

Meanwhile, the ITTF had drafted its document on telematics strategy, "La société européenne face aux technologies de l'information: pour une réponse communautaire," and submitted it to the Council in November 1979. Four of the paper's six points related to telecoms: create a European market for telecommunications and data-processing based on common standards; create a data-services industry competitive in world markets; create a communications network linking EC institutions and national governments; and develop common positions for world telecoms organizations.

The Council requested specific proposals in microelectronics and telecommunications, which the Commission submitted in July 1980. The paper on telecoms strategy argued that new services would be in demand after 1983 and they would require a harmonized, Europe-wide network. A draft recommendation requested the PTTs to open bidding for terminal equipment to all EC manufacturers and to set aside 10 percent of their network-equipment purchases for bidding from any EC maker. The Commission hoped for a Council decision by December 1980 so that the recommendations could be implemented over 1981–83.^[31]

In the event, approval did not come until November 1984. The market-opening proposal obtained approval by not specifying the kind of European company (supplier or maker) it would benefit. The PTTs or other network operators would allow completely open bidding for all new terminals, and (beginning in 1986) for 10 percent of their annual purchases of network equipment and conventional terminals (like handsets, modems, telex).^[32] The measures were weak. For one thing, they were passed in the form of a Recommendation (which is not legally enforceable). For another, the Council approved only an experimental two-year period for the open bidding.^[33]

[30] Agence Europe , 29 December 1979, p. 9.

[31] Agence Europe , 18 July 1980, p. 6.

[32] CEC, Communication from the Commission to the Council on the Status of the Community Telecommunications Policy , 5.

Consequently, as of autumn 1987, though the Germans had diligently listed their calls for tenders in the Official Journal , the Spanish and French had yet to do so.^[34]

In June 1983 the Commission announced that it was preparing a proposal for Community action in telecoms. A document went before the Council at its Stuttgart meeting that month.^[35] It declared that the goal was to stimulate the production and use of the most advanced equipment and services. The Commission was careful to stress that its increasing responsibilities in telecoms would in no way alter the duties of the PTTs nor affect national arrangements for the transfer of PTT revenues to other budgets. The Commission also recognized that the timing was propitious; the Ten needed to cooperate in order to ensure compatibility of new products and services during the period of rapid change due to digitization, satellites, fiber optics, and advances in microelectronics. Finally, the Commission requested the national governments to appoint representatives to a senior officials' group to discuss the establishment of a European telecoms body.^[36]

Unfortunately for the Commission the ITTF was not always as tactful vis-à-vis the PTTs as the Commission had been in the Stuttgart document. Members of the ITTF, who were running the telecoms show within the Commission, spoke of creating a super-PTT at the EC level, to run the EC's networks. Naturally, this proposal enraged the PTTs. One of my informants, who had been a member of the ITTF at the time, said that it made the PTTs suspicious of Commission efforts and reluctant to participate.^[37]

However, the idea of a senior officials' group did take hold. The Industry Council, meeting in November 1983, unanimously agreed with the Commission paper submitted the previous June in Stuttgart and asked the Commission to organize meetings with representatives of the industry ministers and the telecoms industries to prepare an action program by the end of the year.^[38] Davignon and the ITTF responded by creating the Senior Officials Group for Telecommunications (SOGT), composed of a mix of representatives from

[33] Ibid., 5.

[34] Rene de Cazanove, "La quête d'un futur à douze," 53.

[35] "Le CEE souhaite un 'marché commun' des télécommunications," Les Echos , 7 June 1983.

[36] Agence Europe , 17 June 1983, p. 10.

[37] Interview 50; de Cazanove, "La quête," 51.

[38] Agence Europe , 7 November 1983, p. 7.

ministries of industry and economics and from the PTTs. The work of the SOGT provided the programmatic basis and initial political support for Commission telecoms initiatives.^[39]

Meanwhile, at the end of September the Commission had proposed to the Council six lines of action in telecoms. The Commission's arguments in that early document touched on the issues that would remain central to telecoms discussions at the European level. First, European equipment makers needed "home-markets of sufficient size and, under European conditions, these can only be provided by the Community as a whole." The Commission pointed out that nine different switching systems had to try to survive on a fragmented European market, while three systems in Japan and four in the United States each benefited from a large home market. Second, the Commission stressed the threat from the United States and Japan, noting that Japanese industry was oriented by unified planning for the Information Network System and that Japanese exports in advanced telecoms gear were "already now advancing at an astonishing rate in major markets." Third, fragmented planning for ISDN and broadband networks both limited the scale of investment in future networks and services and "threaten[ed] the [competitiveness] of the European telematics industry."^[40]

The six action lines were these:

[39] Interviews 28, 50; CEC, Communication from the Commission to the Council on Telecommunications , COM (84) 277 final, 2.

[40] CEC, Communication from the Commission to the Council on Telecommunications: Lines of Action , COM (83) 573 final, passim.

[41] Ibid., 8–13.

The first three objectives would be addressed in what became the EC's RACE program.

The newly constituted SOGT began a series of meetings with the Commission to discuss possible areas of cooperation. They met six times between November 1983 and March 1984; the consensus that emerged was the basis for Commission proposals for EC telecoms strategy in May 1984. The Commission again emphasized the dangers of U.S. and Japanese competition, stressing the competitive pressure unleashed by deregulation in the United States. Furthermore, the bottom line was that "the capacity to meet these challenges, and to cope in a timely manner with the opportunities born out of the development of telecommunications, is outside the capability of national operators on their own."^[42]

The Commission placed a heavier emphasis than previously on the dynamic economic role played by telecoms. By 1990, the paper stated, telecoms would be the EC's largest economic sector, surpassing even the automobile industry, and accounting for 7 percent of Community GNP by the year 2000. It also noted that telecoms investment had a significant multiplier effect: For every ECU invested in telecoms infrastructure, there was a 1.5 ECU increase in total economic activity. In particular, demand was soaring among business users for advanced services. Thus, telecommunications would create new jobs and save jobs by increasing the competitiveness of firms. In fact, the document declared, telecoms "constitutes a special tool for reviving the economy and protecting employment in the Community."^[43]

The action lines proposed in September 1983 became the outline, after approval by the SOGT, for an EC program in telecommunications. The main elements of the program were:

[42] CEC, Communication from the Commission to the Council on Telecommunications , COM (84) 277 final, 2–3, 10.

[43] Ibid., 3–7.

In announcing the program, Davignon argued that the EC could not be complacent. "The situation in this sector is not bad but it is fragile. Up to the present this is the kind of situation in which the EC reacts the worst, by developing a sense of self-satisfaction," declared Davignon.^[46]

The industry ministers in December 1984 approved the objectives proposed by Davignon. The third goal, development of the technologies needed for future networks and services, took shape in the form of RACE, to which the remainder of the chapter will be devoted.^[47] Before analyzing RACE, however, I will examine the

[44] Type approval refers to the process of verifying that a piece of equipment conforms to standards. The PTTs have historically set the standards and tested for compliance. Mutual recognition of type approvals would thus mean that a terminal approved in one national laboratory would be recognized by other countries as certified for attachment to the network.

[45] CEC, Communication from the Commission to the Council on Telecommunications , COM (84) 277 final, 14–22.

[46] Europolitique , 19 May 1984.

[47] The fourth point, telecoms for the least-favored regions, became the Special Telecommunications Action for Regional Development (STAR) program. Budgeted at 780 MECU (to be matched by participating governments), the program was approved in 1987 to run for five years.

Commission's most recent synthesis of its telecommunications goals and strategies, the 1987 Green Paper.

The Green Paper

Telecommunications intersected with the 1992 project for unifying the internal market, and the result was the Commission's Green Paper. This lengthy document, circulated in May 1987 and released the next month, placed telecommunications reform in the context of revitalizing the European economy. In fact, the actual title of the paper is Towards a Dynamic European Economy . Its first paragraph declares: "A technically advanced, Europe-wide and low-cost telecommunications network will provide an essential infrastructure for improving the [competitiveness] of the European economy, achieving the Internal Market and strengthening Community cohesion—which constitute priority Community goals reaffirmed in the European Single Act."^[48]

The Commission built its argument on the economic importance of advanced telecommunications, the technological changes sweeping the sector, and the reforms already underway in all the member states. The bottom line was that Europe would lose out, both as a competitive source of advanced telecoms equipment and services and as an economic player more broadly, if the changes were not managed jointly at the European level.

Backed by extensive documentation, the Commission concluded the Green Paper with ten positions intended to serve as guides for telecommunications development. I summarize them here.

[48] CEC, Towards a Dynamic European Economy , 9.

In sum, the Green Paper reasserted the Commission's goal to create open EC markets for network equipment, terminal equipment, and telecoms services. A key part of the Commission's approach was open-network provision (ONP), a transnational network operating on the basis of common standards and interfaces. With ONP service providers would have access to the network anywhere in the EC on the basis of established rules of usage and common tariff principles. ONP implied that customers would have access to advanced services from anywhere in the EC and that approved equipment could be attached in any country.

The Green Paper was intended to spark discussion within the Community, and it did. Private associations supported the Commission's plans. The International Telecommunications Users Group (INTUG), representing European national users' groups, strongly endorsed all the Green Paper proposals. The only reservations by INTUG were in areas where the group thought the Commission had

[49] Ibid., Figure 13.

not gone far enough, such as in the preservation of the PTT monopoly on supplying the first telephone set.^[50] The employers' federation and the Roundtable of European Industrialists also seconded the Commission's proposals.^[51] In addition, the Commission carried on intensive discussions with the SOGT and with the national telecoms administrations.

As a result of all the feedback, the Commission prepared a second document, on implementing the Green Paper.^[52] In it, the Commission reviewed the major points of the responses to the Green Paper. It identified areas of broad consensus, areas of general agreement but with criticisms (both from those who thought the proposals in question went too far and from those who argued that they did not go far enough), and areas to be further defined (like satellites). The Commission concluded that a strong consensus supported the major goals of the Green Paper and on that basis set forth a series of specific goals and deadlines, all to be achieved for the completion of the internal market in 1992. The timetable included opening the market for terminal equipment, open competition in services, ONP, common standards, separation of regulatory and operational responsibilities, and other supporting measures.

The Council supported all these objectives in a resolution passed in June 1988. The resolution, which was nonbinding, was without deadlines, but it did request the Commission to propose specific measures to achieve all the goals. In the remainder of this section, I describe the Commission's efforts to implement the Green Paper objectives, and the political rows they have provoked.

One of the central problems in creating a pan-European telecoms network was standards. Standards consist of technical specifications for how electronic information is to be packaged on the network and for how various pieces of network equipment are to talk to each other. Traditionally, the PTTs controlled the standards-setting process in each country and the procedures for verifying that specific pieces of hardware complied with the national standards.

[50] George G. McKendrick, "The INTUG View on the EEC Green Paper," 325–29.

[51] The Roundtable of European Industrialists is an association of about thirty of Europe's largest corporations from both EC and non-EC countries, including Philips, Siemens, Olivetti, Daimler Benz, Volvo, Fiat, Bosch, ASEA, and Ciba-Geigy.

[52] CEC, Towards a Competitive Community-wide Telecommunications Market in 1992—Implementing the Green Paper on the Development of the Common Market for Telecommunications Services and Equipment—State of Discussions and Proposals by the Commission .

Common standards had been one of the Commission's priority themes. In fact, through the initiative of the Commission, in July 1984 the EC signed a memorandum of understanding with CEPT. Under this memorandum, CEPT produces common standards and specifications for equipment approval (type approval, in EC jargon) according to priorities established by the EC. The standards defined by CEPT are called Normes Européennes des Télécommunications (NETs). NETs are the basis for new services and equipment introduced or authorized by the PTTs.

The Commission chalked up another success when the Council, in July 1986, approved a binding Directive on mutual recognition of type approvals (Directive 86/361/EEC). The gist of the Directive was that once a specific type of terminal equipment has been certified as conforming to the relevant common standards, it must be accepted as approved in every other member country. The Directive employed the mutual-recognition approach to standards adopted by the Commission in conjunction with the 1992 program: Rather than suffer through the laborious process of elaborating EC standards, each country must accept as valid approvals granted in any other country. The legal effect is to make NETs mandatory for EC members. The Directive also assigned the Commission the task of preparing each year a list of priorities and a timetable for standardization and common technical specifications. The Commission would then transmit the list to CEPT, which would draft the appropriate NETs.

One priority area has been ISDN, the next step in network evolution. With ISDN, digitized networks will be able to carry simultaneously voice, data, text, and some images. For the range of new services to function across the EC, the ISDNs being created in each member country need to be built to common standards. The Commission proposed that the EC take actions to ensure pan-European ISDN. In the event, the Council passed a (nonbinding) recommendation "on the co-ordinated introduction of ISDN" in December 1986.

In the Green Paper the Commission proposed the creation of a permanent and independent European Telecommunications Standards Institute (ETSI). ETSI came into being in March 1988 despite some conflict over the overlap between its mission and that of CEPT in standards. However, CEPT had certain deficiencies—namely, a lack of permanent, specialized technical expertise, and a lack of representation

for telecommunications industries. In ETSI, network operators (primarily the PTTs), equipment manufacturers, and telecoms users meet to establish technical specifications. Following Commission proposals ETSI produced a timetable for the coordinated introduction of ISDN services in Europe. The PTTs responded by signing a memorandum of understanding in April 1989 on the introduction of pan-European ISDN by 1992.^[53]

The area in which the Commission most vigorously assaulted the old telecoms order was in opening to competition markets in terminal equipment and services. In both terminals and services the Commission sought to liberalize markets by issuing directives under Article 90 of the Treaty of Rome. The general principle of Article 90 is that enterprises to which states grant "special or exclusive rights" must conform to the competition rules of the Treaty. Most important, Article 90 charges the Commission to ensure application of its provisions and grants it the power to issue binding directives to member states. The key feature of such Commission directives is that they do not have to be approved by the European Parliament or by the Council of Ministers. The Commission used Article 90 against the PTTs, arguing that PTT monopolies constituted "abuse of dominant position" (Article 86), which restrains competition. Beginning in 1988 the Commission employed Article 90 to open EC markets in terminals and VANs. All member countries supported the goal of an open terminal market, but several opposed the Commission's unilateral tactic. In fact France (supported by three other countries) challenged the action in the European Court of Justice. The application of Article 90 in services was equally controversial, and several states objected to the goal as well as the tactic. A compromise was struck in December 1989, liberalizing VANs but leaving basic voice telephony and telex as PTT monopolies.^[54]

In short the Commission spearheaded an effort that is creating a new telecommunications regime in Europe, one based on common standards and open markets.

The Politics of Race

The various dimensions of the Commission's telecommunications efforts provide the context without which it would be difficult to

[53] Hans Baur, "Telecommunications and the Unified European Market," 33–34.

[54] See Wayne Sandholtz, "New Europe, New Telecommunications."

understand the role of the RACE program. RACE touches on several of the topics treated in the previous section, like standards and planning future networks and services.

Winning the Support of Industry and the PTTs

As with ESPRIT, the ITTF had prepared itself thoroughly for its assault on Europe's traditional telecoms structures. It had carried out its own assessments of the telecoms sector and had also commissioned a number of technical and economic studies from prominent consultancy firms, including Arthur D. Little International, Mackintosh International, McKinsey, and The Yankee Group. The outside studies provided ITTF members with the facts and figures they needed to be taken seriously. In other words, the studies were not a tool to convince the governments but rather a means of increasing the technical credibility of the Commission. When ITTF officials went before industry groups or the SOGT, they commanded reputable information.^[55]

In fact the Commission began RACE with the same Roundtable of twelve companies that had designed ESPRIT. Why were the Twelve interested in the Commission's schemes for telecoms? The interest in collaborative R&D stemmed largely from the various aspects of technological change examined in Chapter 6. Innovation was dauntingly rapid, and it involved not only traditional telecoms technologies but also VLSI microelectronics, the technologies of dataprocessing and computer networking, and, with an eye to the future, optoelectronics. Collaboration was a way of covering more of the important technological bases than was otherwise possible. But one technology factor deserves special attention in the case of telecommunications: the rising costs of R&D and the attendant imperative for large markets.

The latest generations of public switches were extremely expensive to develop. The costs of developing digital exchanges ranged from $500 million to $1.4 billion. Future optical switches were sure to require substantially more resources. To amortize such large investments in R&D, companies had to achieve vast sales. The Commission cited studies showing that to be economically viable a switch would have to capture 8 percent of the world market. No national

[55] Interviews 28 and 50.

market in Europe surpassed 6 percent of the world total.^[56] Or, as Rob van Tulder and Gerd Junne show, given development costs of the magnitude cited and given that R&D consumes about 7 percent of sales, a switch would have to capture a market worth $14–$16 billion. The difficulty was that no European country had a switch market of that scale. The French market was worth about $10.9 billion, the German one $11.7 billion, and the British one $7.2 billion.^[57]

Because of these pressures, the world telecommunications industry had been consolidating. Data show that telecommunications companies were particularly active in the wave of interfirm alliances that swept the telematics industries. Indeed, a list of interfirm alliances involving telecoms firms since 1980 would consume several pages.^[58] In addition, telecommunications manufacturers were merging at a striking rate. For its equipment business outside the United States, AT&T entered into a joint venture with Philips. Plessey (United Kingdom) bought the smaller company Stromberg-Carlson (United States) and in 1988 formed a joint venture with GEC (United Kingdom), uniting their switching businesses. The biggest merger was the 1986 purchase by CGE (France) of the overseas telecommunications subsidiaries of ITT, forming the world's second largest telecoms-equipment manufacturer, Alcatel. Ericsson (Sweden) subsequently bought the remaining French maker, CGCT. Siemens established a joint venture with GTE (United States). The Italian firm Italtel teamed up with AT&T, and Telettra swapped shares with Spain's Telefonica. Consortia were forming in 1987 to compete for the pan-European mobile-communications market. Alcatel, Nokia (Finland), and AEG (Germany) joined forces, as did Philips and Bosch, and Ericsson had ties with Siemens, Matra (France), and Orbitel (a joint venture of Racal and Plessey in the United Kingdom). By 1987 the telecommunications business was already a highly concentrated one, with the ten largest firms accounting for 67 percent of world production.^[59] The technological and market pressures behind the trend of alliances and concentration made R&D collaboration attractive to European enterprises.

[56] CEC, Towards a Dynamic European Economy , 90.

[57] van Tulder and Junne, European Multinationals , 70.

[58] See Pisano and Teece, Collaborative Arrangements; LAREA/CEREM, Les strategies d'accord; and van Tulder and Junne, European Multinationals , 60–63.

[59] Jacques Arlandis, "Le dilemme des quarante fabricants," 65.

The Commission had decided that starting anything in telecommunications would require action from two major groups: industry and the PTTs. Politics required that the Commission approach the PTTs first. The major companies active in telecoms equipment depended on close working relations with their PTTs and on PTT orders. For political reasons, therefore, they could not meet with the Commission until after the PTTs had done so. Shortly after the creation of the SOGT in late 1983, the Commission proposed meetings with the research arms of the PTTs (like the Centre National d'Etudes des Télécommunications in France or the Forschungsinstitut der Deutschen Bundespost in Germany). The telecoms administrations were extremely reluctant and ended up sending only low-level representatives.^[60]

However, now the Commission could initiate meetings with industry. Representatives of the ITTF visited the leadership of the Roundtable companies, laid out the Commission's analysis of the challenges, and proposed that industry help in working up an R&D program aimed at broadband communications. The companies accepted. The result was the Planning Exercise in Telecommunications (PET). Over two and a half months in the summer of 1984 eighty experts from industry and the PTT research organizations met twice a week to determine the R&D activities that could benefit from Community-level cooperation. The nitty-gritty work took place in four technical panels organized by industry. In looking over the list of participants, I could identify only two national telecoms laboratories represented on the technical panels: Nederlandse PTT and the telecoms research center for the IRI-STET group.

The result of the PET was a volume of over 600 pages titled Proposal for an Action Plan: RACE . The document outlined a vision for broadband systems and proposed research areas and tasks to prepare for them. The overall objective, by consensus of the PTT labs and industry, was "Community-wide introduction of Integrated Broadband Communication (IBC) by 1995 taking into account the evolving . . . ISDNs." The IBC network envisioned here would eventually include all national and local networks and would subsume all the services presently offered on distinct networks, plus ISDN, cable television, and mobile communications. It would involve fiber-optic transmission down to and possibly including the

[60] Interviews 28 and 50.

customer premises. IBC would unite in one network all business and residential services, from telephony to data transmission to massmarket HDTV. The proposed RACE program would provide the technological base for IBC.^[61]

The body of the document detailed the R&D work to be carried out in RACE and proposed a separate program of IBC demonstrations and trials. The PET experts offered a timetable that included dates for the various phases of RACE, the demonstrators, and installation of the working system. They aimed for initial commercial availability by 1995. The RACE work would require an estimated 5,196 man-hours over five years. The R&D was divided into four areas:

A fifth area, not detailed in the report, would be low-cost, high-performance terminal equipment. For the four research areas, the document specified 115 separate tasks. For each task, objectives, technical approach, connections with other tasks, milestones, a time scale, and the resources needed (in man-years) were specified. The PET recommendations became the basis for the Commission's later proposals for the RACE program.

The Commission distributed the PET report to the PTTs for review. This was the hook that eventually brought in the telecoms administrations. The PTT directors saw that concrete, credible technology planning was going on with the heavy involvement of industry. It became clear, first, that if they were not involved, the PTTs would lose control of telecommunications development, and, second, that by participating the PTTs could inject their will into the proceedings. In fact one of my Commission informants told me that a PTT official vented her displeasure with events but conceded that the Commission had them "ensnared."^[62] PTT participation became

[61] CEC, Proposal for an Action Plan: RACE , 0.1.1–0.1.4.

[62] Interview 28.

important later in the process, for although the decision to approve RACE was a political one and although various ministries (economics, industry, research) had input to the governments on the question, the PTTs were politically central.

Even though the PTTs were beginning to warm to the idea of RACE, the Commission ran into a temporary setback, largely of its own making. In late 1984 the Commission proposed to the SOGT a 1 BECU RACE program that was flatly rejected. The proposal was shot down so quickly that it never appeared in the press. Some in the ITTF had wanted to put a telecoms program on the books before Davignon left the Commission at the end of 1984. After that failure the Commission regrouped and drew up a smaller Definition Phase.

The Definition Phase

By March 1985 the Commission had put together a proposal for a preliminary, limited Definition Phase to precede the full-fledged RACE program. At the time the Commission foresaw a ten-year RACE program with a total EC contribution of about $400 million. Planners in Brussels hoped that the Definition Phase could be approved before the summer, so that it could issue a call for proposals in the autumn.^[63]

With the proposal for the RACE Definition Phase, the Commission submitted to the Council a report on the R&D requirements in telecommunications. The report reiterated the importance of telecommunications to the European economies, especially as the convergence of telecommunications, data-processing, and audiovisual media proceeded. It also noted that R&D costs for telecoms equipment were soaring at the same time that the life expectancy of a public exchange was falling—ten to fifteen years for an analog exchange, less than ten years for a digital switch. A collaborative program would help ease some of the resulting pressures. The Commission argued further that the need for joint action was urgent given the efforts of Europe's competitors.^[64]

The report also reaffirmed the consensus reached by the PTT research arms and industry in the PET—namely, that the objective

[63] Europolitique , 16 March 1985.

[64] CEC, Report of the Commission to the Council on R&D Requirements in Telecommunications Technologies as Contribution to the Preparation of the R&D Programme: RACE , 19–21.

should be an EC-wide transition to IBC. Toward that goal a collaborative R&D program should commence. The program's basic aims would include production of a reference model for IBC (its networks, terminals, and services), development of the technological foundation for that IBC, and close cooperation between RACE and existing telecoms standards organizations. The document also reproduced the timetable proposed in the PET—a Definition Phase (1985–86), a main Phase (1987–92), and a final phase (1992–97), with installation of the IBC network beginning in 1992 and operation in 1995.^[65]

The proposal specified that the R&D work would be precompetitive (as in ESPRIT), executed via contracts between the Commission and EC organizations. The total EC contribution to the Definition Phase would be 22.1 MECU over eighteen months; the EC would cover half the costs of projects and participants would put up the other 50 percent. The proposal included a ceiling of 4.5 percent of the EC contribution that could be allocated to Commission staff.

The Definition Phase work plan encompassed two main parts, each subdivided into specific areas:

[65] Ibid., 24–26.

For the three areas under Part I objectives, approach, methods, and tasks were specified. The areas under Part II also contained a number of specific tasks.^[66]

The Commission's proposals for a Definition Phase were taken up by the Council in early June. The ministers of industry and of telecommunications, meeting jointly, did not arrive at any unanimous position on RACE. The three large countries expressed reservations about the level of funding proposed (22.1 MECU from the EC) and also argued that work on defining a reference model for IBC should be carried out in the association of PTTs, CEPT.^[67]

The research ministers were more receptive than the industry and telecommunications ministers to the Definition Phase, although with reservations. The United Kingdom and Germany emphasized that a decision about the Definition Phase would not imply any commitment to a later Main Phase. Further, the research ministers stipulated that the Definition Phase work would comprise forecasts and estimates of R&D needs in the area, rather than actual R&D. Britain and Germany wanted the Definition Phase funding to come out of the existing Commission budget for R&D. The Commission ended up assenting to these conditions.^[68]

The Research Council was able to reach a general political agreement on the RACE Definition Phase after a compromise was reached on the ties between the EC program and CEPT. The research ministers noted that the ministers responsible for telecommunications desired that CEPT play a central role in defining the IBC reference model. France in particular clung to that view, and Britain and Germany favored it as well. In essence, this was a move to ensure that the PTTs, through CEPT, would retain some control over planning for the future system. The director-general of the French PTT, Jacques Dondoux, speaking later in June at a meeting of CEPT, articulated that view: "A telecommunications Europe cannot be made only at the level of the EEC and industry; the network operators, who alone

[66] CEC, Proposal for a Council Decision on a Preparatory Action for a Community Research and Development Programme in the Field of Telecommunications Technologies: R&D in Advanced Communications-Technologies for Europe (RACE), Definition Phase , passim.

[67] Agence Europe , 5 June 1985, p. 5; Alain Bradfer, "Project RACE: Les européens sont bien changeants," Le Matin , 5 June 1985.

[68] Agence Europe , 5 June 1985, p. 5; Paul Cheeseright, "Setback for EEC Telecommunications Plan," Financial Times , 4 June 1985, p. 2; and Quentin Peel, "Broad Agreement on Ambitious EEC Plan for Communications," Financial Times , 5 June 1985, p. 2.

can ensure the interconnection of national networks, should play a greater role."^[69]

The Commission wanted to retain control of the entire program, pointing out that CEPT included sixteen non-EC countries, that it had extremely tenuous relations with industry, and that it was organizationally and financially inadequate for the job. But facing the big three states, the Commission was forced to compromise, agreeing that the preparation of the IBC reference model would take place in close collaboration with the PTT association. The French research minister, Curien, declared himself satisfied that CEPT would make a strong contribution to the Definition Phase.^[70]

The Research Council, however, could not pass a final decision because of a dispute involving a separate technology issue, the proposed EC laboratory for the handling of tritium, which is essential for research on nuclear fusion. The Italians withheld final approval of the Definition Phase in order to obtain approval for establishing the tritium laboratory at the Joint Research Center at Ispra in northern Italy. Only the French opposed the Ispra site. By the end of July the French had withdrawn their objection to the Ispra laboratory, and the Council of Ministers gave final approval to the RACE Definition Phase on 25 July 1985.^[71]

As passed by the Council, the Definition Phase received a budget of 20 MECU, of which 14 MECU would go to the Part II projects open to bidding. The remaining funds were assigned to Part I (the IBC reference model). Part I was subject to restricted tendering because it was entrusted to CEPT and the PTTs. Incidentally, CEPT was not equipped to perform the reference-model work. In order to fulfill its role in RACE, CEPT created the first fixed structure in its history—a permanent secretariat, to be complemented by groups of experts hired for full time.^[72]

The Council Decision authorizing the RACE Definition Phase defined operating procedures that had been settled in ESPRIT. Part II

[69] Guillaume Goubert, "La longue marche des télécommunications vers la normalisation," La Tribune , 2 July 1985.

[70] "RACE sur les rails," La Tribune , 6 June 1985; Alain Bradfer, "Project RACE," Le Matin , 5 June 1985.

[71] Philippe Lemaitre, "Les Dix confirment leur volonté," Le Monde , 6 June 1985; CEC, Background Information for the Call for Tenders for the RACE Definition Phase , 1.

[72] Guillaume Goubert, "La longue marche . . . ," La Tribune , 2 July 1985.

projects had to involve at least two independent industrial partners from different member states. Participants would pay half the cost of each project, and the work would be carried out within the EC. The Commission would be responsible for managing the program in close cooperation with a Management Committee composed of two representatives from each member state. The Management Committee would have to approve any Part I contracts worth over 100,000 ECU and any Part II contracts over 400,000 ECU.^[73]

The proposal and selection procedures mirrored those for ESPRIT. Consortia of companies, universities, PTT laboratories, and other research organizations submitted proposals. Commission administrators, assisted by experts hired for that purpose, evaluated the proposals in several steps. First, they assessed the adequacy of the managerial arrangements and the fit of the technical content to the work plan. At this stage the proposals were anonymous. Next, the same set of evaluators looked at the qualifications of the proposers and their personnel. Finally, the evaluators examined the funding request. The proposals passed through two or three different sets of evaluators. The Commission then submitted a list to the Management Committee, recommending some proposals for quick approval, some for additional scrutiny, and some for rejection.^[74]

After the research ministers had given general political approval to the Definition Phase, the Commission published in the Official Journal in April 1985 an "advance notice," inviting expressions of interest. Following final Council approval in July, the official call for tenders attracted more than eighty proposals involving 171 EC organizations. The Commission brought in about seventy experts from industry and academia for a period of two weeks to assess the proposals. In February 1986 the Commission announced the thirty projects that received funding; they would run until the end of 1986.^[75] The Commission also contracted for special consulting studies on HDTV with several organizations. The broadband network would have the bit-rate capacity to carry HDTV signals. The HDTV studies reflected the emerging consensus that the extension

[73] Council Decision of 25 July 1985, 85/372/EEC, Official Journal , no. L/210, 7 August 1985, pp. 24–26.

[74] Interview 9; CEC, Background Information for the Call for Tenders for the RACE Definition Phase , passim.

[75] Interview 9.

of broadband networks into residences would be economically feasible only if they could offer entertainment services with potential for mass markets—like HDTV.

The largest of the Definition Phase projects was one to define a subscriber-premises reference model. It was headed by GEC of the United Kingdom, involved twenty-nine organizations, and received 3 MECU. The breakdown of national representation on the projects is presented in Table 8.2 European subsidiaries of five American-based multinationals figured among the winning consortia. In fact ITT affiliates were involved in thirteen of the thirty-one projects—more than any other company. Eight of the ITT participations were by Standard Elektrik Lorenz, the West German subsidiary. Other American companies represented in the Definition Phase included AT&T, GTE, Hewlett-Packard, and IBM. The participation of companies with U.S. ties disturbed some of the European firms, but

neither the Commission nor the Council made an issue of it. The only stipulation in the rules governing the Definition Phase was that the R&D be performed within the EC.^[76]

The RACE Definition Phase projects concluded at the end of 1986. Part I activities produced an initial IBC reference model via the efforts of the PTTs through the permanent nucleus of experts (the Groupe Spécial Large Bande) established within CEPT. An industry group advised the PTT-CEPT group working on the reference model. The model developed general outlines of the switching and transmission capabilities of the future IBC network and the kinds of services it could be expected to support. The services envisioned were grouped into four categories as shown in Table 8.3. The referencemodel work also came up with three possible plans for evolution toward IBC. The first was based on circuit-switching techniques (in which connections are made by establishing a circuit that utilizes all the users' available channels). The second assumed asynchronous time division (in which a stream of "packetized" information of all kinds—voice, image, data—travels through a single channel, routed by the network switches). The third plan was based on the assumption of a fully optical network. Further clarification of the eventual services and research into the technological options would permit choices to be made among the plans.^[77] The part of the Definition Phase open to tenders from EC companies, universities, and labs was also extremely successful.

Commission officials expressed a high degree of satisfaction with the Definition Phase and considered that it had legitimized RACE.^[78] A survey of participants, including industry, universities, and network operators, revealed strongly positive reactions to the program. When asked whether the Definition Phase approach (exploring the technologies and the functional specifications simultaneously) was appropriate for European objectives and conditions, 85 percent answered affirmatively. The question "Do you consider that your participation in the [RACE Definition Phase] has

[76] Richard L. Hudson, "Many in Europe Decry Inclusion of US Firms in EC Phone Project," Wall Street Journal , 13 February 1986; Richard L. Hudson, "EC Commission Gives 109 Grants on RACE Study," Wall Street Journal , 26 February 1986; Ivo Dawnay, "US Groups Prominent in Community Project," Financial Times , 26 February 1986, p. 2. By 1987 ITT's telecommunications subsidiaries in Europe had been purchased by France's CGE, forming Alcatel.

[77] CEC, RACE: Consolidated Preliminary Report of the RDP Projects , 2-1–2-9; B. Catania, The Many Ways Towards IBC , 19–20.

[78] Interviews 28 and 50.

contributed significantly to a clarification of the issues and helped you in your own work?" elicited a positive response from almost 90 percent of respondents. Over 90 percent of respondents answered yes to the query "Has working directly together with your colleagues from other countries been of benefit and do you consider

this to be important to the strengthening of Europe's posture?"^[79]

The Main Phase

Before the Definition Phase had reached its conclusion, the Commission was already putting the final touches on its proposal for a Main Phase of RACE. Like ESPRIT II, RACE was packaged in the 1987–91 Framework Programme on R&D. RACE Main therefore became stuck in the political quagmire surrounding the Framework Programme, described in the previous chapter. Thus, although the Commission proposal went to the Council in late October 1986, it would be a year before the program was finally approved.

The Commission document proposing RACE Main rephrased slightly the overall goal that had been stated in the Definition Phase: "Introduction of . . . IBC taking into account the evolving ISDN and national introduction strategies, progressing to Community-wide services by 1995."^[80] The change in the statement is the phrase "and national introduction strategies," reflecting increased sensitivity to the diverse plans and levels of advancement of the national telecoms administrations.

The proposal was for a five-year program with an EC contribution of 800 MECU, bringing the total program (with cost sharing) to 1,600 MECU and 10,000 man-years of work. The Commission raised the possibility of a second five-year phase but made clear that its present proposal in no way implied or required one. It argued that the Definition Phase had already demonstrated the need for a coordinated European approach to IBC development and that rapid follow-on work was needed to maintain momentum.

The Main Phase would include three principal parts:

[79] CEC, Proposal for a Council Regulation on a Community Action in the Field of Telecommunications Technologies: RACE , 20–21.

[80] Ibid., 5.

The Commission hoped for a launch date of 1 January 1987.

Because of the logjam holding up the Framework Programme, nothing happened on RACE through the first half of 1987. One Commission official considered the delay fortunate, as it allowed the ITTF (by then the Information Technologies and Telecommunications Task Force) time to digest the results of the Definition Phase and incorporate them into the first RACE work plan.^[82] The draft work plan was finished in the spring and published in June 1987.

The work under Part I would focus on the continual updating and refinement of the reference model. In addition, it would include the mapping of the reference-model functions onto alternative implementation strategies. Each strategy would be broken down into a chronological sequence of frames, each frame representing a credible real situation. Each strategy was thus a model of the evolution, in concrete steps, toward IBC. The work plan contained two principal strategies.

Part II aimed at developing the technologies needed to implement the evolution strategies. The R&D therefore would pursue diverse technical options. The work in Part III would focus on two areas: the development of verification tools, and IBC applications pilot schemes (an addition). The pilot projects would ideally be transnational experiments to try out various IBC functions and equipment.

All three parts included a number of specific tasks, some ninety-nine in all, that proposals would be expected to address. Each task description included sections on background, objective, technical

[81] Ibid., 28–30.

[82] Interview 50.

approach, and key results and milestones. The work for Part I and the verification aspect of Part III would, as in the Definition Phase, center on the permanent broadband nucleus established in CEPT. The rest would be open to bidding.

As the Main Phase proposal sat on the shelf, it became increasingly clear that RACE had won broad support, both in industry and among the PTTs. The PTTs had discovered that they could exercise significant influence within RACE and had to maintain a role in the process that was underway. Indeed, it is striking that after the Definition Phase the PTTs did not oppose RACE. Commission officials noted that the problems in winning approval for RACE Main stemmed not from the national telecoms administrations but from the Framework Programme debates. One RACE administrator said that by the time he started circulating the draft work plan for the Main Phase, he had high-level contacts in the PTTs and was well received by them.^[83]

Industry also lent strong support to the Commission's efforts in telecommunications. In October 1986, as the Commission was preparing its Main Phase proposal, the Roundtable of European Industrialists issued a paper on telecoms in Europe from the point of view of major business users. The Roundtable paper highlighted the problems caused for European businesses by diverse national equipment-approval rules, the lack of compatibility of services across borders, and the absence of coordinated planning for new services and networks. The paper summarized the Commission's initiatives (including coordinated ISDN, standardization, open markets, and RACE) and stated strong support for them: "Among users, the will exists to bring the EEC's objectives to fruition." The Roundtable recommendations paralleled the Commission's own.^[84]

In January 1987 the industrial association Union des Industries de la Communauté Européenne (UNICE) submitted to the Commission a set of proposals for urgent action in the sector. It was the first public effort involving both makers and users of telecoms equipment to influence telecommunications policy at the EC level. The UNICE proposals paralleled the themes stressed by the Commission since 1983: an open European market for terminals, open public procurement, common standards, an end to PTT monopolies,

[83] Interviews 28 and 50.

[84] Roundtable of European Industrialists, Clearing the Lines: A Users' View on Business Communications in Europe , 16–19.

and continued development of EC R&D programs. The industrialists stressed the importance of precompetitive research.^[85] All these themes emerged in the Commission's Green Paper, which circulated a few months later (as discussed previously in this chapter).

The Community telecoms industry enthusiastically backed the proposed RACE Main Phase. The only significant exception was Siemens, which, according to industry and EC sources, was reticent about the program.^[86] Siemens executives explained that they were definitely in favor of strong European standards and that cooperative R&D was all right as long as it remained precompetitive. However, Siemens thought it was ahead of RACE objectives and would possess a competitive advantage in the future.^[87] Even so, Siemens ended up participating in five of the first batch of Main Phase projects. In general, though, RACE had solid backing from industry.

The Commission had published in the Official Journal of 5 December 1986 a call for "expressions of interest" in RACE Main. The following March the ITTF distributed a book listing the names and addresses of the nearly 600 respondents (some organizations submitted multiple responses, in different areas).^[88] In June the draft work plan was released, and on 1 July 1987 the Commission published in the Official Journal a "call for a reserve list of tenders." The call took the form of a "reserve list" because the Framework Programme—and with it RACE—had not been passed by the Council. But the Commission wanted to be able to move quickly once the anticipated approval finally came. By the October deadline, the Commission had in hand ninety-five proposals, involving 320 organizations.^[89]

As detailed in the previous chapter, the member states finally reached a political agreement on the Framework Programme in July 1987, with formal ratification in late September. The component

[85] Marc Paoloni, "Les P et T contre l'Europe?" La Tribune , 24 January 1987; William Dawkins, "Industry Outlines Telecoms Policy for EEC," Financial Times , 23 January 1987, p. 3.

[86] Interviews 50 and 60; Philippe Lemaitre, "La Commission européenne veut consacrer 5,5 milliards de francs aux télécommunications," Le Monde , 21 October 1986.

[87] Interview 45.

[88] CEC, RACE: Expressions of Interest .

[89] Interview 9; CEC, RACE '88: The RACE Programme in 1988 , 1.

parts of the Framework Programme had to be approved by the Council individually; this approval came for RACE on 14 December 1987, with Community funding of 550 MECU (down from the 800 MECU originally hoped for by the Commission).^[90]

The administration of the Main Phase follows the pattern set by ESPRIT and the Definition Phase. The RACE Management Committee must approve the yearly work plans and the list of projects chosen by the Commission. The prime contractor for each project must report monthly (in summary fashion) on technical and financial progress. In addition, each project is subject to an annual review, with the results to aid in updating the work plan and also possibly in revising the projects themselves. The Commission also provided for the project managers of all RACE projects to meet jointly two or three times a year to ensure communication among projects. The entire program was to be assessed after thirty months of operation and again at the completion of the five years. The Regulation establishing RACE left open the possibility of a second five-year phase.^[91]

The September ratification allowed the Commission to proceed with the evaluation of the proposals and negotiation of contracts. The evaluation of the proposals followed the same criteria and procedures as in the Definition Phase and utilized the services of seventy-three outside experts.^[92]

In the end forty-six projects were retained, with an EC contribution of 186 MECU over three years.^[93] The first batch of projects covered most of the tasks in Parts I and II of the work plan. Work began in January 1988. As in ESPRIT II, countries of the EFTA were permitted to participate as full members of the consortia, though they could not receive EC funds. Of course, through Part I work managed by CEPT, all twenty-six CEPT members had an indirect hand in RACE.

With the first year's projects underway, the Commission began the first annual revision of the work plan. The process built on contributions from and consultations with some 2,000 parties. The

[90] CEC, RACE '88 , 1; CEC, European Telecoms Fact Sheet 6: The RACE Programme .

[91] Official Journal , no. L/16, 21 January 1988, p. 35.

[92] CEC, Background Information for the "Call for a Reserve List of Tenders" for the RACE Programme: General Information , 9–11.

[93] CEC, RACE Workplan '89 , v.

1989 work plan became the basis for a second call for proposals, which was issued in June 1988. The second round of proposals was to focus especially on Part III of the work plan, functional integration—the demonstration and testing of the technologies needed to support and integrate in a single network the services envisioned in the reference model. With the second round of contracts, eighty-four RACE projects were underway. Table 8.4 shows the levels of national participation in RACE through the second round. RACE

projects involve 362 organizations; of the 230 EC companies participating, 90 are small firms.^[94]

Although no one can predict the full range of future IBC applications, they will probably be composed of a limited number of basic functions, also called service primitives. The work plan for 1989 included tasks aimed at developing concepts and tools for verifying that emerging IBC components and subsystems (and eventually entire systems) perform the needed basic functions. To this end a number of new tasks were added to Part III, including several applications pilots that will test prototype IBC services for specific classes of users. The areas to be addressed in the applications pilots include:

Work on the applications pilots will interact with the other parts of the program to refine the model and the technologies.^[95]

Conclusion

Given the situation in European telecommunications circa 1980, few could have predicted the extent of collaboration that emerged over the next eight years. At the beginning of the decade, national PTTs held a monopoly not only on telecommunications networks, equipment, and services, but also on policy-making. By 1988 that was no longer the case. Because of Commission leadership, EC states

[94] CEC, RACE '89 , 1–3.

[95] CEC, RACE Workplan '89 , passim; and CEC, Background Information for a "2nd Call for Proposals" for the RACE Programme: General Information, , passim.

now pursue joint actions in planning future networks and in R&D. They have taken important collaborative steps in standards, type approval, and the opening of markets for equipment and services. In effect the Commission is at the center of an emerging new regime for telecommunications in Europe; it forced CEPT to assume new roles and created ETSI.

The theoretical framework of this study proves useful in explaining the cooperation. First, on the demand side, purely national strategies of the traditional European kind were discredited. Technological change and deregulation abroad were undermining the PTT monopolies, and governments increasingly realized that the nationalistic fragmentation of the past would not serve their present and future needs. Thus, policy-makers were in the adaptive mode, searching for new approaches to the telecommunications sector; major studies and reforms were in progress in all the EC countries by the mid-1980s.

Second, political leadership was crucial in organizing cooperation. The perceived inadequacy of national strategies and the adaptation undertaken by governments created an opportunity for policy leadership on the part of the Commission. It seized the initiative and convinced the governments and the PTTs that cooperation was essential. The same transnational industrial coalition that brought forth ESPRIT allied itself with the Commission and lobbied powerfully for telecommunications collaboration.

Finally, RACE benefited from organizational arrangements that provided for a satisfactory distribution of the benefits. The program is big enough to offer a large number of pieces of the action. Additionally, as in ESPRIT, states participate in a modified version of à la carte: They contribute to the whole program but participate only in parts that their domestic organizations find attractive.

For all the debate and confusion surrounding telecoms liberalization in the United States and Japan, the process in those countries was simple compared with what transpired in Europe. European states not only began to liberalize but also acted to end decades of nationalistic fragmentation in telecommunications. The results constitute a significant political achievement.