Document Sample
					Sectoral Systems in Europe - Innovation, Competitiveness and Growth

      Under the Fourth Research and Technological Framework Programme,
                  Targeted Socio-Economic Research, TSER
               [Contract no: SOE1-CT 98-1116 (DG 12-SOLS)]

                Work Package 2: Sectoral Systems in Europe
               RP3 Telecommunication hardware and services

        (global system for mobile telecommunications)

                         Working Paper ESSY

         Leif Hommen with the assistance of Esa Manninen

            Department of Technology and Social Change
                  Linköping University, Sweden
GSM (Global System for Mobile Telecommunications):


                       Leif Hommen

         Department of Technology and Social Change
                    Linköping University
                     S-581 83 Linköping
                     tel: +46 13 28 44 52
                    fax: +46 13 28 44 61

                    with the assistance of

                         Esa Manninen
         Department of Technology and Social Change
                    Linköping University
                     S-581 83 Linköping
                     tel: +46 13 28 29 68
                    fax: +46 13 28 44 61

                 Report prepared for
               the research network on
        European Sectoral Systems of Innovation

                      30 April 2001


GSM (Global System for Mobile Telecommunications): ................................................... 1
Contents............................................................................................................................... 3
Introduction ......................................................................................................................... 5
1. Historical and Geographical Profile................................................................................ 7
   1.1. Development Period: 1980s - 90s .......................................................................... 7
   1.2. Year of Introduction: 1992 (with a second stage in 1995-1996)......................... 7
   1.3. Geographical Coverage: Regional/International (Europe; Overseas) .............. 8
   1.4. Provenance: European Consortium of PTOs, Suppliers, and SDOs................. 8
2. Technical Characteristics .............................................................................................. 10
   2.1. Equipment............................................................................................................. 10
      2.1.1. Terminals......................................................................................................... 11
      2.1.2. Access Network - Radio Base Stations ........................................................... 12
      2.1.3. Switching......................................................................................................... 12
   2.2. Technical Capabilities.......................................................................................... 13
      2.2.1. Terminals......................................................................................................... 13
      2.2.2. Access Network - Radio Base Stations ........................................................... 13
      2.2.3. Switching......................................................................................................... 14
   2.3. Functions ............................................................................................................... 14
      2.3.1. Mobile Voice Telephony................................................................................. 15
      2.3.2. Data Transmission........................................................................................... 15
      2.3.3. Security............................................................................................................ 16
   2.4. Services.................................................................................................................. 17
3. Related Systems and Standards..................................................................................... 19
   3.1. Complementary Standards.................................................................................. 19
      3.1.1. Cordless Telephony: From Telepoint to DECT .............................................. 19
      3.1.2. Cellular Telephony: PCN and DCS1800 ........................................................ 21
      3.1.3. Paging Systems: ERMES ................................................................................ 24
      3.1.4. Trunked Mobile Radio/Data:TETRA.............................................................. 25
   3.2. Rival Standards .................................................................................................... 26
      3.2.1. D-AMPS.......................................................................................................... 27
      3.2.2. CDMA............................................................................................................. 29
      3.2.3. PDC ................................................................................................................. 31
   3.3. Summary ............................................................................................................... 32
4. Public Sector Actors...................................................................................................... 35
   4.1. European National PTOs or PTTs and SDOs ................................................... 35
      4.1.1. The Changing Role of PTOs/PTTs in Europe................................................. 36
      4.1.2. The Swedish Case: Televerket ........................................................................ 37
      4.2. European SDOs (CEPT and ETSI) ................................................................ 39
      4.2.1. CEPT (Conférence Européenne des Administrations des Postes et
      Télécommunications) ................................................................................................ 39
      4.2.2. ETSI (European Telecommunications Standards Institute) ............................ 41
      4.2.3. The MoU Group .............................................................................................. 43

   4.3. Other Public-Sector Actors ................................................................................. 46
      4.3.1. Public Research Organisations........................................................................ 46
5. Private-Sector Actors .................................................................................................... 49
   5.1. Equipment Suppliers ........................................................................................... 49
      5.1.1. Ericsson ........................................................................................................... 50
   5.2. Service Providers & Network Operators ........................................................... 56
      5.2.1. A New Policy Regime in the European Union ............................................... 56
      5.2.2. ... and in Sweden ............................................................................................. 58
      5.2.3. Telia Mobitel ................................................................................................... 59
      5.2.4. Comviq ............................................................................................................ 61
      5.2.5. Europolitan ...................................................................................................... 62
   5.3. Other Private-Sector Actors................................................................................ 63
      5.3.1. GEAB .............................................................................................................. 64
      5.3.2. Talkline............................................................................................................ 64
6. Main Outcomes ............................................................................................................. 65
   6.1. Consequences: European Competitive Advantage; Nordic Dominance......... 65
   6.2. Changes to Charging Systems............................................................................. 67
   6.3. Growth Statistics .................................................................................................. 70
      6.3.1. Growth Figures (Equipment) .......................................................................... 70
      6.3.1. Growth Figures (Services) .............................................................................. 72
7. Conclusions ................................................................................................................... 75
   7.1. Knowledge Base & Learning Processes ............................................................. 75
   7.2. Firms, Non-Firm Organisations & Networks.................................................... 77
   7.3. Geographical Boundaries .................................................................................... 78
   7.4. Long-Term Dynamics of the Sector and Co-Evolutionary Processes ............. 81
   7..5. Public Policy......................................................................................................... 83
   7.6. European International Performance and Comparisons with US & Japan .. 84
References ......................................................................................................................... 86


The topic of this report is the development of the Global System for Mobile
Telecommunications (GSM) standard, which was the major European contribution to the
development of the second generation of mobile telecommunications technology. It
forms part of a larger research project, undertaken in collaboration with the research
network on European Sectoral Systems of Innovation (ESSY).

The larger research project on mobile telecommunications examines the evolution of
mobile telephone equipment (systems) and services since the introduction of the NMT
(Nordic Mobile Telecommunications System) 450. To describe and analyse the
development of mobile telephony during the last few decades, it uses technical standards
as an ordering device. Thus, the project starts with the first Nordic mobile telephone
standard, NMT450, and later standards are successively introduced as they have emerged
over time up to the present day. These standards - in the order: NMT450, NMT900,
GSM, and UMTS/MBS - have been chosen because they represent a continuous
progression through the three generations of mobile telephony.

Using the NMT standard as a starting point leads naturally to a focus on Sweden and the
other Nordic countries. Parallel standards in other geographical areas during the early
days of mobile telephony will be discussed to some extent, but the main focus is on
actors and events in Sweden. As the second and third generation digital mobile telephone
standards are introduced, the geographical coverage is expanded, but the case of Sweden
is still used to illustrate specific phenomena under investigation.

This report, as already mentioned, deals primarily with the second-generation standard,
GSM. GSM is, of course, placed in its historical context, and therefore some references
are made to the preceding first-generation standards, such as NMT. However an equal, if
not greater amount of attention is devoted to other standards of the same generation.
Similarly developments in Sweden are discussed, not only with reference to the historical
background but also in relation to the wider context of contemporary developments
within the European Community. The structure of the report is the following:

Historical and Geographical Profile

This section provides a short background on how the standard was developed, its
development period, and when it was introduced on the market. The geographical
coverage and organisational provenance of installed systems is also reported.

Technical Characteristics

A brief description of the main technical characteristics is given. The different kinds of
equipment involved and their relations within the system, as well as services offered in
    This section is adapted from a text originally authored by Esa Manninen.

connection with the system, are discussed. Concerning mobile telephone services offered
in connection with each standard, it should be noted that the services are described
mainly from an end-user’s perspective.

Related Systems and Standards

The discussion of systems and standards related to GSM focuses on systems and
standards of the same generation. It distinguishes two main kinds of ‘relations’. The first
of these is complementarity; the second, rivalry. Complementary systems and standards
include a number of European innovations in digital mobile telecommunications that
were introduced in connection with GSM, usually under the auspices of ETSI. Rival
systems and standards refer to non-European second-generation technologies for mobile
telecommunications that were conceived as direct competitors to GSM.


Two separate sections under this heading discuss key actors involved in the creation of a
specific standard, but also actors active in the industrial, mass-market phase of the
process. The examination of actors and their relations is intended to capture the
organisational and institutional changes that the creation of the new technical standard
entailed. A more detailed analysis of key actors also makes it possible to study main
features of the knowledge-base related to the standard, and how the knowledge-base
changed over time.

The first section discusses public-sector actors (PTTs/PTOs, SDOs, universities, public
research organisations). The second section discusses private-sector actors
(telecommunications equipment producers, private service providers). Both of these
sections devote considerable attention to the relations among actors. Hence, the relations
of public sector actors with private sector actors are discussed, and vice versa. Earlier
sections of this report distinguish between equipment and services. Thus, the discussion
of actors also distinguishes between equipment producers and service producers (which
in turn may be separated into network operators and service providers).

Main Outcomes

The purpose of this section is to capture the degree of commercial success of the new
standard, including both the equipment side and the services side of the industry. The
information that is reported in this section includes figures on number of subscribers,
number of terminals sold, market shares for the different mobile telephone operators, etc.

Findings and Conclusions

Finally, some tentative findings and conclusions are presented. These are stated in
relation to some general questions addressed by the ESSY research network. It should be
noted, however, that these findings and conclusions are only preliminary statements,
based on work that is still in progress.

1. Historical and Geographical Profile

1.1. Development Period: 1980s - 90s

Work towards the definition of the GSM standard was first initiated in 1982 by CEPT
(Conférence Européenne des Administrations des Postes et Télécommunications). This
organisation represented the PTOs and PTTs of 25 European countries and was then the
principal body in Europe responsible for the development of telecommunications
standards. At the urging of some of its members, primarily those from the Nordic
countries and the Netherlands, ”CEPT formed a new standards group -- Groupe Spécial
Mobile (GSM) -- with the mandate to specify a new radio telephone system for Europe”
(Garrard 1998: 126).

The impetus for developing GSM arose from the widely perceived need for a pan-
European standard in mobile telecommunications. A wide variety of first-generation
systems was implemented in Europe during the 1980s. Consequently, most users could
not roam to more than a few other countries. Moreover, the market for equipment had
been fragmented. There was also remarkable growth in first-generation mobile
telecommunications. It soon became evident, though, that the development of new
systems with greater subscriber capacity would be necessary to sustain continued
expansion of mobile telecommunications networks. New systems were also required to
meet demands for new service capabilities, such as more extensive ‘roaming’. But, at the
same time, the lack of market co-ordination provided insufficient incentive for the
development of such systems.

1.2. Year of Introduction: 1992 (with a second stage in 1995-1996)

In 1985 and 1986, the Groupe Spécial Mobile drafted a list of recommendations for the
GSM standard, and tests were conducted on eight prototype systems submitted by
different consortia (Lindmark 1995: 111). The results of the tests, carried out in France
by the national telecommunications research laboratory, CNET, were presented at a
plenary session of GSM held in Madeira in 1987 (Garrard 1998: 129).

During the same period, the European Commission became increasingly involved with
the GSM standardisation process, leading in 1987 to the formulation of a Directive and
Recommendation by the council of Ministers. In the same year, a Memorandum of
Understanding (MoU) was drafted by the original participants of Groupe Spécial Mobile
and signed by telecommunications operators and regulators from thirteen countries. The
MoU committed the signatories to introduce GSM by the 1st of January, 1991, though
this date was later postponed by six months (Garrard 1998: 131).

Manufacturers and operators carried out tests to validate the GSM Recommendations in
1989 and 1990. These tests resulted in the discovery of technical problems that led to a
further six months or more of delay in the introduction of GSM. The standard was
eventually launched, with the initiation of commercial GSM services in 15 European
countries, in 1992. However, a lack of ready terminal equipment at that time meant that
there were ”no mobiles to sell” (Garrard 1998: 161 - 164).

A second phase of the GSM standard was introduced in 1995-1996. By this time, the
involvement of the European Commission had led to a gradual hand-over of the
responsibilities for the standardisation of GSM (and other European telecommunications
standards) from CEPT to the recently established ETSI (European Telecommunications
Standards Institute) (Garrard 1998: 134 - 136). The transfer of responsibilities was
accompanied by a migration of the Groupe Spécial Mobile (now the Special Mobile
Group, or SMG) from CEPT to a new organisational base in ETSI.

1.3. Geographical Coverage: Regional/International (Europe; Overseas)

GSM, as noted above, was conceived from the outset as a pan-European standard. It was
thus at first a regional/international standard, in terms of both origin and coverage.
However, GSM has increasingly spread to countries and regions outside of Europe, with
the result that GSM, in some respects, at least, has verged on becoming a ‘world’ system.

By September, 1996, signatories to the GSM MoU included 167 operators from 103
countries and states, there were 20 further applicants from participating countries, and 10
new applicants from previously non-participating countries (Garrard 1998: 164). The
non-European regions represented included Africa, Asia, Australasia and the Middle East
(Garrard 1998: Table 5.9). ”To the surprise of many Europeans, GSM had even made
inroads into North America where a number of newly licensed PCS operators in the USA
and Canada chose to use GSM technology in their allocated spectrum band at 1900 MHz
(Garrard 1998: 164).

1.4. Provenance: European Consortium of PTOs, Suppliers, and SDOs

As explained above, GSM is essentially a regionally-based ‘consortium’ standard, which
was finally ratified and fully formalised by the EU-based European Telecommunications
Standards Institute, ETSI, although it had its origins in another European standards
development organisation, CEPT (Conférence Européenne des Administrations des
Postes et Télécommunications). The main force behind this organisational shift was the
European Commission, which during the mid and late 1980s became increasingly
involved in backing and promoting the GSM standard. The entire process of developing
the standard was led by the Groupe Spécial Mobile, which was later renamed the Special
Mobile Group (SMG) after completing its migration from CEPT to ETSI.

Historically, monopolistic national PTOs have been the dominant actors in both CEPT
and ETSI, although equipment manufacturers have also become increasingly important to
their activities in standard setting. (Moreover, other operators have gained some
influence, now that the PTOs’ former monopolies are no longer protected). GSM was an
important milestone in this respect. In the early stages of developing the GSM standard,
as in past practice, manufacturers participated in the process only by invitation and not
‘by right’. As the development of the GSM standard progressed, however, it became
increasingly apparent that the co-operation of manufacturers was vital to the success of so
large and complex a project. Thus, when ETSI was formed, with the agreement of CEPT
members in 1988, membership was opened to any European organisation involved in
telecommunications (Temple 1992: 177). This was in keeping with the principle that
future telecommunications standards in Europe should be developed by a forum ”that
would allow, as a right, contributions from manufacturers, users, research bodies and
private network operators” (Garrard 1998: 133 - 134).2

Within this organisational framework, a number of sub-consortia were formed and
competed under the umbrella of the broader GSM consortium. For the most part, these
were alliances of PTOs/PTTs and major equipment manufacturers, often with strong
national ties, that collaborated on competing technological designs for the GSM system.
The several alliances formed within the GSM consortium are described in some detail in
sections 4 and 5 below.

  This author does not provide a clear definition or specification of the term, ‘users’. However, he uses the
term primarily in reference to individual end-users of the services provided through telecommunications
networks (Garrard 1998: 145 - 157, 469 - 470). Another source, more specifically focused on issues of
user-involvement in the development of telecommunications standards, differentiates between individual
end-users and commercial ‘intermediate-users’ (Hawkins 1995: 23). The former type of user can usually
be defined in the traditional sense of a ‘subscriber’ -- ”an entity utilising and/or extending telephony based
facilities but not primarily involved in providing such facilities to other parties” (ibid.: 23). The latter type,
however, is more amorphous. The ‘intermediate-user’ is likely to be involved in some form of ”third-party
service provision” and to have ”acquired a measure of control over the addition of value to basic public
network facilities” (ibid.: 23).

2. Technical Characteristics

The GSM system was a digital system, which meant that the voice transmission between
the terminal and the base station would be digitalised. Digitalisation would have
profound effects on the development of mobile telecommunications technology. GSM,
like other digital systems, could accommodate a much larger number of subscribers than
analogue systems such as NMT, and it was therefore capable of overcoming the ‘capacity
problem’ that had imposed a limit on the further development of analogue systems for
mobile telecommunication. In addition, GSM was an ‘open’ standard -- one that allowed
producers to configure communications between the system’s components in different
ways, thus shifting the responsibility for system configuration from network operators to
equipment producers. An important competitive consequence for the producer firm was
that it ”either ha[d] to be a system provider or have alliances with others” (McKelvey,
Texier, and Alm 1998: section 2.3).

2.1. Equipment

GSM was, as mentioned above, a digital system. For most users, this meant primarily
that their terminal equipment had greater capabilities -- for example, with respect to
roaming -- and more functions. But digitalisation also meant that network operation
would require the build-up of an entirely new infrastructure, with implications for the
division of labour among key actors.

For terminal equipment, digitalisation meant not only improvements in the quality of
voice transmission and enhanced security functions, but also the capability to receive
transmissions of data, as in so-called ‘short messaging’. The introduction of a separate
subscriber identification module (SIM) both enhanced security and accessibility for users
and made possible a wide range of new services. Some of them involved actors in
sectors such as financial services that formerly had little or no involvement with mobile

Infrastructural developments also implied new arrangements for and among the key
actors. In the GSM access system, radio base stations became ‘smarter’ at the expense of
switches. GSM radio base stations from different manufacturers were, as a result, highly
compatible and could be easily substituted for one another. However, they also became
much more complex, requiring much more software and programming. This implied a
reconfiguration of the knowledge base of producing firms. Developments with respect to
switching both presumed and made possible increased co-operation among network
operators. By means of complimentary ‘home’ and ‘visitor’ location registries, network
operators could effectively track roaming subscribers and extract payment from them.

2.1.1. Terminals

The GSM terminals offered further improved voice quality, even though not comparable
to that of the fixed network. Furthermore, bugging of an ongoing call was made more
difficult because of encryption of the voice signals (Meurling and Jeans, 1994: 170). In
addition, a much wider variety of terminal equipment, capable of meeting a growing
diversity of user requirements, became available with the advent of GSM and other
digital standards.

Originally, five classes of terminals, ranging from extremely high-powered devices to
very low-powered ones, were included in the GSM standard. However, the basic choice
for most users turned out to be between vehicle-mounted mobile phones (usually Class 2)
and portable hand-sets (usually Class 4). Portable hand-sets have been by far the most
common choice. However, a great variety of options soon emerged, even for this one
type of device. By 1994 there were 18 different models of Class 4 hand-sets available
from 11 manufacturers, as well as basic models and new prototypes available from
original equipment manufacturers or OEMs. (Garrard 1998: 150 - 151)

In general, the functionality of GSM hand-sets was greatly increased compared to
analogue hand-sets. One important means by which this was made possible was the
separation of the subscriber’s identity and the mobile telephone, in order to improve
access and security for users. This was accomplished by introducing a separate electronic
Subscriber Identity Module (SIM) that could be plugged into different terminals by the
same subscriber. The SIM’s capacity for information extended well beyond the
subscriber’s identity, so that it could record a wide variety of details about personal
communications preferences and could also be used to store received short text messages
in the subscriber’s absence. Thus, users could in principle develop fairly detailed service
profiles, and new services could be offered. For example, the SIM’s compatibility with
financial cards created new possibilities, not only for the billing and payment of mobile
telephone services (and hence the reduction of administrative costs for network operators
and service providers), but also for the delivery of other financial services via mobile
telephones. (Garrard 1998: 151 - 153)

Later developments in hand-set technology included the dual-band hand-set developed by
Ericsson to operate on GSM networks operating in both the 900 MHz and 1800 MHz
radio frequency bands (Holst 1997: 108). Another innovation, Short Messaging Services
(SMS), was a feature of the GSM standard from the beginning. However, SMS was at
first not very well developed beyond the basic capability for subscribers to receive short
text messages of up to 160 characters, which could be stored by the SIM and displayed
by a mobile terminal (Garrard 1998: 156). Since it was first demonstrated by Nokia in
1994, SMS has undergone considerable development. (See subsections 2.2.1., 2.2.2.,
2.3.2., and section 2.4.) For example, the original transmission rate of 9.6 kb/s was
improved to one of 14.4 kb/s by 1998 (Emmerson 1998: 40). At that time, however,
serious problems still remained with the connection of SMS applications to existing
information systems, most of which had different proprietary interfaces. ‘Average’ users,

moreover, continued to find the SMS interfaces in mobile phones to be ”awkward”.
(Pohajakallio 1998: 48)

2.1.2. Access Network - Radio Base Stations

GSM radio base stations were designed to be more ‘intelligent’ than their analogue
predecessors. GSM radio base stations incorporated a control unit that was
proportionally much more complex and important than its counterpart in, say, the NMT
system. Comparatively, the control unit in GSM radio base stations required much more
software development and computer programming, and the development of the access
system in general required a much broader competency base. (McKelvey, Texier, and
Alm 1998: section 2.3)

A noteworthy feature of the radio access system in GSM was the economic advantage
conferred on network operators by the use of multiple voice channels on a single radio
‘carrier’. In brief, this feature made it possible for the network operators to achieve
considerable economies of scale. Although GSM was an extremely complex system,
with very high development costs, the associated infrastructural costs turned out to be
lower than in analogue systems. In contrast to the latter, GSM did not require a special,
‘dedicated’ base station transceiver for each channel used. Instead, GSM radio base
stations required only one-eighth the number of transceivers required to achieve a similar
level of carrying capacity. They were also capable of doubling this capacity by adding
only one more eight-channel radio module at a relatively low cost, compared to total base
station cost. Thus, average GSM investment costs per individual business subscriber were
initially only three-quarters of those for mature analogue networks, and by the mid-90s it
was projected that they would soon fall to nearly one-third of analogue investment costs
per subscriber. (Garrard 1998: 158)

2.1.3. Switching

The decision to concentrate network intelligence in the radio base stations, rather than in
switches, taken by the Groupe Spécial Mobile in 1988, ensured the inter-operability of
switches and radio-subsystems. This design choice would also drive down infrastructural
costs. It was particularly intended to make the switches produced by different suppliers
interchangeable, thus providing operators with competitive sources of supply for
switches. A similar logic had been followed in the NMT system, and it was therefore
somewhat ironic that Ericsson, one of the main suppliers of switches for NMT, opposed
its further application in the case of GSM. However, such opposition was over-ruled.
(Garrard 1998: 144)

The most notable initial developments with respect to switching in the GSM system were
the advances that were made in order to achieve the objective of an extended
international (originally, pan-European) roaming capability. The register that enabled
subscribers to be traced and calls routed to them in earlier cellular systems became, in
GSM, divided into two components: the Home Location Register (HLR) and the Visitor
Location Register (VLR). The HLR could be interrogated by the operators of other GSM

systems than the ‘home’ system, in order to verify the identities and authorisation of
subscribers originating from that system and ‘roaming’ into theirs. The VLR provided
complimentary functions for ‘home’ systems, and also recorded information about calls
made on the visited network. Co-operating network operators could thus jointly monitor
and bill their roaming subscribers. (Garrard 1998: 144)

2.2. Technical Capabilities

Some basic capabilities of the GSM system were mentioned above, in section 2.1. In the
following paragraphs, some additional details of a more technical nature are discussed.
The discussion also attempts to link some of the main features of the initial version of
GSM technology to more recent developments. However, it is possible to offer only a
brief overview here, not a detailed exposition.

2.2.1. Terminals

Some basic capabilities of GSM terminal equipment have already been identified. These
include improved voice quality, greater functionality, the storage and display of non-
voice data transmissions, and dual-band capabilities. Speech-coding in the GSM system
actually made possible a greater reliability of voice quality. It remained slightly lower
than in a perfect analogue transmission but could be sustained in almost any successful
GSM transmission (Garrard 1998: 147 - 148). SIMs, in addition to providing greater
security and access to users, also made possible the reception and display of data
transmissions. The data capability of GSM terminals was originally limited to the
shortest of ‘short messages’. However, their capacity to receive and download data has
been under constant improvement since 1994, when Nokia introduced one of the first
data interface products -- a ‘PC’ card that can be inserted into a portable computer
connected to a mobile hand-set by means of a cable (Garrard 1998: 148 - 149). Dual-
band hand-sets were a natural accompaniment to the development of new GSM networks,
such as DCS1800, which used different bands of the radio spectrum. (See section 3.1.2.
for a discussion of DCS 1800.) In addition, dual-band hand-sets facilitated more
extensive global ‘roaming’ among GSM networks, by now established in different
continents as well as different countries (Johnston 1998: 53).

2.2.2. Access Network - Radio Base Stations

One of the outstanding technical features of the GSM system was its greatly enhanced
capacity. In contrast to analogue systems, in which each frequency transmitted
corresponded to only one traffic channel, the initial GSM system had eight traffic
channels and could handle both speech and data transmissions. Time-division
multiplexing provided multiple time-slots for the traffic channels. Moreover, the later
development of half-rate voice-encoding would double the number of traffic channels for
each carrier from 8 to 16. (Garrard 1998: 144 - 145) The original data transmission rate
of GSM networks was 9.6 kb/s. By 1998, though, the development of high speed circuit-
switched data (HSCSD) made possible a 14.4 kb/s service with GSM. The further

development of HSCSD was expected at that time to double this increase several times
over, eventually reaching 57.6 kb/s data rates (Pohajakallio 1998: 42). More recently, the
development of general packet radio service (GPRS) -- a solution that (like HSCSD)
combines channels and employs new coding techniques, but (unlike HSCSD) does not
depend on continuous connection time -- has made possible ”data transmission at rates
ranging from 14kb/s to 115 kb/s and higher” (Clever 1999: 41).

2.2.3. Switching

Since switches, as opposed to other components, had already been digitalised prior to the
advent of GSM, switching technology was not a primary target for development in GSM.
Instead, ”the access system [became] the focus of attention, in the hunt for new and more
cost-effective technology to connect the subscribers” (Meurling and Jeans 1995: 243). In
this connection, however, the interface between the radio access system and the switching
system acquired an ever-increasing importance in GSM. The development of capabilities
for data transmission has been particularly important in this respect. In particular, a
progression from SMS to later, more ‘comprehensive’ solutions such as HSCSD and
GPRS (discussed above) has required the development of specialised complements to
switching and base station control systems. With GPRS, for example, the minimal
requirements of a packet data network (PDN) include a serving GPRS support node
(SGSN), a gateway GPRS support node (GGSN), and a ”dedicated frame relay network -
- or connections through the mobile switching centres (MSC) -- link[ing] base station
systems to the GPRS serving nodes” (Clever 1999: 44). More complicated solutions
may be required under higher traffic conditions.

These and other developments in GSM -- for example, the emergence of high-capacity
PCNs (personal communications networks) discussed below, in subsection 3.1.2. -- have
greatly increased the exchange of signalling information that must be processed by the
control systems in network exchanges. For that reason there has been an ongoing effort
to increase the capacity of processors connected to the switching system. In the mid-
1990s, for example, Ericsson was engaged in a project intended to achieve a ten-fold
increase in current processor capacity. It would result in a call-handling capacity
between 240 and 300 times that of the original APZ processor for the AXE switching
system (Meurling and Jeans 1995: 253).

2.3. Functions

The foregoing accounts of equipment (section 2.1) and the technical capabilities of
various components in the GSM system (section 2.2) provide a basis for discussing some
of the main functions of the GSM system. Thus far, such discussion has been limited
mainly to outlining some key implications for manufacturers, network operators, and
service providers. Here, the main focus will be on specific facilities offered by the GSM
system, considered from the end-user’s perspective.

From what has been noted, it is clear that digitalisation made possible the development of
a number of specialised services in GSM that analogue systems such as NMT had not
been capable of offering. These will be discussed in section 2.4. At this point, however,
the main concern is only with three general functions: mobile voice telephony (including
issues of both quality and mobility); data transmission; and security.

2.3.1. Mobile Voice Telephony

The main function of GSM, for most users, has been mobile voice telephony. There was,
as observed earlier, an overall improvement of voice quality in the GSM system, due to
new speech-coding techniques. As shown in preceding discussions of terminal
equipment, there was also considerable progress in hand-set technology. Hand-sets
became smaller and more lightweight, and dual-band capabilities were introduced,
without compromising the power of hand-sets. Portability was improved; hence,
mobility was enhanced.

Extensive international roaming capabilities were also introduced with GSM, and this
also enhanced mobility. ”Roaming is now available, not just within Europe but is also
widely used in the Asia Pacific region. Even roaming between these regions is possible”.
(Garrard 1998: 155)

2.3.2. Data Transmission

At the time when GSM and other digital systems in mobile telephony first appeared, they
confronted competing technologies for mobile data communication. Mobile telephony’s
main competitors in this field were specially designed mobile data networks, which
continued their operation into the 1990s. In 1995, Lindmark and Granstrand observed
that “Several mobile data networks have been competing internationally, supported by
groupings of suppliers and operators, thus being, in our terminology, a case of systems
competition” (Lindmark and Granstrand 1995: 381). These authors state that the most
successful of these had been a Swedish system. “MOBITEX, jointly developed by
Ericsson and Telia and internationally supported by RAM Mobile and Bell South, has
been successful in this competition, now being the most widespread mobile data
communications standard on the international market” (ibid.: n. 3).

However, mobile data networks eventually proved to be an inferior solution compared to
wireless data transmission via digital mobile telephony. The competition between these
two types of system is reviewed at some length by Garrard (1998: Chapter 11, section 3),
who also comments extensively on the case of MOBITEX (ibid.: 430 – 431). This author
points to the poor growth record of MOBITEX, even in its ‘home market’, Sweden, prior
to the advent of GSM and other digital mobile systems. He also draws the following
conclusions concerning the disappointing performance of MOBITEX and similar systems
in relation to their target market:

           The failure has sometimes been attributed to the lack of standards, or to the
           limited performance of available technology, but the real reasons seem to be

            exactly the same as those identified by Televerket in Sweden in 1991 – the
            effort required to share a limited-capacity data channel between many users is
            just too great. That does not mean, however, that mobile data will not be an
            increasingly important aspect of mobile communications. (Garrard 1998:

Analogue mobile telephony systems like NMT were not readily capable of transmitting
data reliably and affordably (Paetsch 1993: 28). However, digital systems possessed the
basic capability for data transmission. In GSM, this was first introduced in the
rudimentary form of SMS, or Short Messaging Services. Since then, however, there has
been a rapid evolution in capabilities for data transmission, as discussed in the preceding
two sections. (See, in particular, subsections 2.1.1., 2.2.1. and 2.2.2.)

SMS originally emerged as a functional equivalent to paging systems, which were a
necessary complement to analogue mobile telephones that had no capability to receive
and store messages. However, the superior coverage provided by paging systems has
meant that ”SMS is unlikely to provide a substitute” (Garrard 1998: 156). On the other
hand, the two-way data exchange capability of SMS has shown great potential for further
development. It can provide Internet access through HTML-compatibility, now being
developed in the form of the wireless application protocol, or WAP (Pohajakallio 1998).

2.3.3. Security

Security, privacy, and confidentiality were problematic in analogue systems like NMT.
Call interception and the ‘cloning’ of mobile telephone numbers became common forms
of fraud. In the US, particularly, ”Fraudulent use of roamer mobile identification
numbers (MIN), reaching levels of 25% to 35%, posed a serious threat to the cellular
operators in the mid-1980s” (Paetsch 1993: 159).

In GSM, as in other digital systems, encryption of the voice signals militated against the
bugging of ongoing calls (Meurling and Jeans, 1994: 170). The A5 algorithm that is used
for encryption ”is almost impossible to break, ensuring complete confidentiality for
users” (Garrard 1998: 154). In addition, the introduction in GSM of a separate electronic
Subscriber Identity Module (SIM) that could be plugged into different terminals by the
same subscriber improved not only the accessibility of the system but also the security of
its users (Garrard 1998: 151 - 153). For example, the SIM would protect owner of a
mobile phone from fraudulent use of the terminal if it were stolen. Generally, user
security has been one of the strongest features of GSM from its inception:

            The security measures built into GSM do not simply rely on the complexity
            of the transmission patterns but are explicit features of the standard. They
            provide all the security features that users could require -- anonymity,
            protected authorisation, and confidentialuity. Anonymity is achieved by the
            fact that the only time a subscriber ‘s real identity is sent over the air is when
            his mobile telephone is switched on. The system then allocates a temporary
            identity that is used until the telephone is switched off again. If an

            unauthorised listener wanted to intercept the calls of a specific subscriber, it
            would be impossible even if he had been able to solve the problem of
            decoding the transmission. (Garrard 1998: 153 - 154)

2.4. Services

One reason to introduce the digital standard GSM (Global System for Mobile
Communication) was that it could handle a much larger number of subscribers than the
older analogue standards (McKelvey, Texier, and Alm 1998: 16). This was an important
feature of the new system, because the number of subscribers grew much faster than
predicted in the early years of cellular telephony.

International roaming has been, from the outset, one of the key services in the GSM
system. The GSM operators have made roaming agreements with each other in Europe,
but also with GSM operators outside Europe. For example, the British operator
Vodaphone made agreements with the Australian GSM operators (Meurling and Jeans,
1994: 125).

Such agreements, particularly within the European Union, were actively encouraged by
the European Commission. The EC, in its 1994 Green Paper on a Common Approach in
the Field of Mobile and Personal Telecommunications, took the position that “roaming
agreements should be promoted” (Commission of the European Communities 1994: 29).
It stated that “the provision of services in the context of such roaming agreements
represents the exercise of the freedom to provide services in a Member State other than
the one in which the Service Provider is established” (ibid.). Moreover, the EC was
explicit in this document that “Such activity should not be subject to any regulatory
restriction” (ibid.).

At the same time that the EC made these statements, however, some analysts pointed out
that GSM, as the spearhead of EC efforts to internationalise competition among
telecommunications operators and service providers in Europe, was likely to create some
unprecedented regulatory problems. “In principle”, one observed, “the system allows for
the import and export of services, by using cards from all over Europe” (Pisjak 1994:
299). Although roaming agreements might be concluded by specific operators and service
providers, there was no regulatory framework to govern the roaming contracts. Instead,
network operators were left to decide over the trade in mobile telephony services. In this
respect, it was noted, “the technical system challenges regulation” (ibid.).

The development of capabilities for data transmission in the GSM system has been
discussed in some detail in previous sections. (See, in particular, subsections 2.1.1., 2.2.1,
2.2.2., and 2.3.2. ) Here, the additional observation can be made that the possibilities of
greatly increasing the extent of the GSM system’s capacity for data transmission through
solutions such as HSCSD (high speed circuit-switched data) and GPRS (general packet
radio service) have held out the promise of developing a ”mass market for mobile data”

(Clever 1999). One observer of user response to the recent evolution of data-
transmission within the GSM system has, for example, made the following comment:

            The use of data services is expected to radically change the way a mobile is
            used. Already, subscribers are using e-mail on their laptops connected to
            mobile services. By 1998/99, ... the general packet radio service (GPRS) ...
            will make multimedia services and high-speed Internet access possible”.
            (Shankar 1997: 44)

The introduction of the subscriber identity module (SIM), as discussed above in section
2.1.1., made possible not only better security for GSM subscribers, but also new
possibilities for the payment and billing of mobile telephone services. Among other
things, this innovation enabled network operators and service providers to reduce their
administrative costs (Garrard 1998: 151 - 153). Moreover, the possibilities for cost
reduction in such areas were also accompanied by the creation of greater possibilities for
competition that should ultimately benefit users by ensuring that costs would be lowered,
where possible.

The SIM not only afforded users greater flexibility with respect to payment and billing
through the introduction of innovations such as the pre-paid card. It also introduced a
separation between service subscriptions and specific items of terminal equipment,
making it possible for users to switch either or both service providers and terminal
equipment with relative ease. Moreover, the European Commission, from an early stage
in the development of GSM, remained committed to the principle that using additional
means to lock terminals to any one network was an ‘anti-competitive’ practice (Garrard
1998: 246).

All of these aspects of GSM as a second generation technology for mobile
telecommunications had significant implications for the development and provision of
services. However, the most important aspect of GSM and other second generation
technologies in relation to services was the complete ‘digitalisation’ (or ‘digitisation’) of
mobile telephony. This development not only made transmission and switching systems
more efficient but “also allowed the decoupling of the transmission and processing
functions, formerly integrated, thus opening the possibility of decoupling services
provision from transmission” (Pisjak 1994: 292). Digitalisation facilitated the entry of
many new services (and new providers of both new and existing services) into the
market. Ultimately, it raised “the regulatory question of which services, if any, could or
should be reserved for the operator of the telecommunications infrastructure” (ibid.).

3. Related Systems and Standards

In discussing systems and standards related to GSM, it is possible to distinguish two main
kinds of ‘relations’. The first of these is complementarity; the second, rivalry.
Complementary systems and standards include a number of European innovations in
digital mobile telecommunications that were introduced in connection with GSM, usually
under the auspices of ETSI. Rival systems and standards refer to non-European second-
generation technologies for mobile telecommunications that were conceived as direct
competitors to GSM.

3.1. Complementary Standards

Most of the complementary systems and standards that were introduced during the 1980s
were ‘new service concepts’ intended to broaden the market penetration of digital mobile
telephony by providing adaptations specially suited to particular market segments.
However, not all of these new service concepts were based on cellular radio systems. It
is therefore important to make a technical distinction between ‘cordless’ systems
designed to serve very small geographical areas at very low power levels and ‘cellular’
systems that used low-powered terminals within networks comprised of small cells
(Hawkins 1995: 31 - 32). Both types of system were pioneered in the UK, though
variants were soon introduced in other European countries, including Sweden.

In addition to complementary standards for mobile telephony, it is also possible to
distinguish complementary standards for technologies that are themselves complements –
or, at least not direct competitors -- to mobile telephony. To the extent that such
technologies do not provide complete substitutes, they can instead be considered to form
complements (Lindmark and Granstrand 1995: 381). A clear case of this type of
complementarity can be found in paging systems, which were discussed above (in sub-
section 2.3.2) as continuing to provide a necessary complement to digital mobile
telephones, due to their capacity to provide superior coverage (Garrard 1998: 156). The
development of public access trunked mobile radio/data systems is another such case.
Standards complementary to GSM have been developed by ETSI for each of these
complementary technologies.

3.1.1. Cordless Telephony: From Telepoint to DECT

‘Telepoint’, a technology developed in the UK, was one of the first digital ‘cordless’
systems. It permitted cordless terminals to be connected to the public network if they
were used within a few metres’ radius of a Telepoint base station. Although the fact that
users could not receive calls but only make them led Telepoint to fare poorly on the
market, it was nevertheless adopted by ETSI in 1991 as a provisional cordless standard
re-named ‘CT2’ (FINTECH 1991). At the same time, ETSI declined to recognise the
technically superior ‘call and receive’ cordless system, CT3. CT3, originally developed

by Ericsson, was, however, widely favoured within the telecommunications industry
(FINTECH 1990).

ETSI’s decisions regarding the CT2 and CT3 systems concluded a long debate, which
began with an initial rejection by the EC of CT2. The initial UK application to approve
an EC standard for Telepoint had incited a backlash, with intense lobbying from other
countries in favour of CT3. Prior to the inauguration of ETSI, the opposition to CT2,
combined with reluctance to endorse any ‘single-country’ system as the basis for a pan-
European standard in digital cordless telephony, had led the EC to delay repeatedly its
approval of a CT2 standard for Telepoint. In the meantime, however, the UK
government had continued its support for the technology, despite the lack of a common
standard for all UK manufacturers.

Eventually, a Common Air Interface (CAI) agreement was concluded with European
manufacturers and PTOs/PTTs, the EC embargo on the sale of CT2 products in the UK
was lifted, and CT2 was launched -- first in the UK, then in Europe, and finally overseas.
However, CT2 enjoyed only modest success in the UK and elsewhere. One reason for its
limited and uneven market diffusion was the lengthy (ten-year) time-scale needed to
bring CT2 to the market. But there were also other reasons, such as insufficient
infrastructure, high cost compared to existing analogue mobile systems, and competition
from other digital cordless and cellular systems appearing on the market at roughly the
same time. (Garrard 1998: 447 - 456)

The main cordless competitor to CT2 was the ‘call and receive’ DECT (Digital European
Cordless Telephony) standard, developed in ETSI during the initial launch period of CT2.
DECT, which was approved by ETS in 1992 and released on the market in 1993, was the
product of collaborative work jointly funded by the EU and European manufacturers of
telecommunications equipment. In addition, it was officially supported and promoted by
the EU, through both ”policy statements ... and co-ordinated implementation programmes
for digital wireless telephony” (Hawkins 1995: 31).

DECT shared many technical characteristics with CT2, including the use of an air-
interface specification, but it also drew upon some key design features of the rival CT3
technology, such as ”the time-division techniques of the Ericsson system” (Garrard 1998:
449). Moreover, DECT was technically superior to CT2 in certain respects. Not only
was it a ‘call and receive’, rather than ‘call only’ system, it could also handle a greater
volume of traffic, and required much less infrastructure in the form of fixed base stations.
In addition to its superiority to early versions of CT2, DECT also received strong backing
from the European Commission, ETSI and the largest European PBX (private branch
exchange) manufacturers (Siemens, Alcatel, Ericsson, and Phillips) (Paetsch 1993: 308 -

For both technical and other reasons, therefore, DECT became an important basis for the
development of WPBXs (wireless private branch exchanges) in Europe (ibid.).
Moreover, its capability to provide two-way data services within buildings made it,
according to ETSI, the key technological platform for the provision of on-site wireless

data exchange within Europe (Microcell 1991: 6). However, DECT’s performance limits
with respect to data transmission (initially low bit-rates of less than 10 Mbps), ongoing
competition from other systems (including later versions of CT2), and challenges from
regulatory agencies (such as CEPT) and EU member states (such as Germany and the
UK) all hampered its further development (Paetsch 1993: 314). ETSI announced its
intentions, at the outset of the 1990s, to use DECT as the foundation for developing
WLANs (wireless local area networks) capable of high-speed (20 Mbps) data
transmission (Microcell 1991: 6). For the reasons cited here, though, these plans met with
scepticism from some observers and analysts (Paetsch 1993: 314).

DECT’s development within ETSI was advantageous, ensuring its compatibility with
GSM standards. Thus, ”it could be used as an alternative means of access to any GSM
network” (Garrard 1998: 457). Nevertheless, the DECT system was both difficult to
implement, due to its complexity, and costly compared to other cordless systems.
Consequently, it did not diffuse successfully to the mass market, remaining confined,
instead, to the niche market it was originally intended for -- business PABXs (Private
Automatic Branch Exchanges) -- without having any implications for direct network
service. Towards the end of the 1990s, the market diffusion of DECT was still limited
mainly to this segment and it appeared ”unlikely that DECT will be adopted outside
Europe” (Garrard 1998: 457).

In spite of the initial difficulties and ongoing scepticism by industry analysts concerning
the standard, ETSI proceeded with further development of DECT. In 1996, it created an
“autonomous Project within ETSI” with an ongoing mandate for the development and
maintenance of standards on DECT (ETSI 2001a: 1). The Institute now presents DECT
as a commercial and technological success. In 2001, ETSI was able to report that DECT
had evolved into a complete system for providing “Wireless Local Loop”, and that
“frequency bands have been made available for DECT in more than 100 countries”
(ibid.). It further stated that dual-mode DECT/GSM handsets had become widely
available, and that a new CTM (cordless terminal mobility) Access Profile, allowing
users “to roam between public and private networks”, had been implemented in Italy in
1998 (ibid.). In addition, ETSI also touted DECT as a promising ‘third generation’

            ETSI DECT applied to be one of the technologies [in] the IMT-2000
            specification. The DECT standard is widely supported as the only cordless
            technology among those proposed, and the only IMT-2000 technology that is
            available today. ITU confirmed that DECT fulfils all the necessary
            requirements. (ETSI 2001a: 1)

3.1.2. Cellular Telephony: PCN and DCS1800

By the time that DECT was launched on the market, in 1993, it encountered competition
from an alternative ETSI technology referred to as PCN (Personal Communications
Network). PCN, implemented in 1993 under the DCS1800 standard approved by ETSI in
1991, was a digital mobile telephone system that combined both ‘cellular’ and ‘cordless’

design features (Hawkins 1995: 31; Law 1991: 14). (The acronym, DCS, stood for
‘digital cellular system’.)

Like Telepoint, the PCN concept originated in the UK in the late 1980s, in response to
the Department of Trade and Industry’s (DTI’s) efforts to procure and implement mobile
telecommunications systems that could bring mobile telephony into the mass market
(Garrard 1998: 171 - 176). As a ‘cellular’ solution, PCN was intended to use low-
powered terminals and small cells, attempting to achieve costs comparable to those of
fixed telephone systems (Becker 1991; Ramsdale 1991). As a direct competitor to
DECT, PCN was a ‘micro-cell’ application of GSM and therefore possessed a complete
network structure, rather than being an ‘add-on’ (Haddon 1991).

The UK’s PCN initiative managed to avoid many of the problems that Telepoint had
encountered in relation to the EC. In this case, licenses were initially awarded under an
ETSI-approved standard in 1991, and ETSI subsequently undertook the co-ordination of
the two years of development required to produce a technology that met the
specifications of DCS1800. Consequently, a satisfactory level of market co-ordination
was achieved, at least among producers and network operators at a comparatively early
point in the development of PCN. Thus, ”the UK PCN licensees decided to develop a
common technical infrastructure -- the ‘parallel network architecture’ or PNA -- and to
compete on the basis of services and prices rather than network facilities” (Hawkins
1995: 32; Ramsdale 1991).

Despite the relative ease with which standardisation was accomplished and the
advantages that this conferred on PCN, it also created problems for the holders of
licenses for PCN network operation. These were problems, however, whose solution
would contribute to the successful diffusion of GSM to the mass market. This connection
was due to the fact that ETSI’s adoption of the DCS1800 standard had aligned PCN very
closely with GSM. The alignment was, in fact, so close that ”From a purely technical
point of view, there is absolutely no difference between the services and facilities that can
be provided by GSM and DCS 1800 ... although DCS1800 brings added technical
difficulties” (Garrard 1998: 183).

There were several such technical problems. First, PCN lacked any obvious technical
differentiation from regular GSM technology. For example, both types of system were
configured for lightweight ‘hand-set’ terminals, though there were other options in GSM.
Second, PCN hand-sets were supposed to be sold at lower prices, even though the
terminal equipment in both systems had to be equally complex and capable of the same
processing power, as well as being similarly light-weight. This was especially difficult,
due to the fact that the DCS1800 standard allocated a radio frequency for PCN that was
twice as high as that allocated for GSM. Because components for radio transmission at
this frequency level were not at the same advanced stage of manufacturing as those for
the lower GSM level, PCN hand-sets were at first more costly to produce than those for
GSM. Third, the DCS1800 frequency allocation for PCN dictated a much higher density
of radio base stations -- about 4 to 6 times as many as for GSM at its minimal density --

in order to provide adequate coverage. Thus, network infrastructure costs of PCN were
also higher than those of GSM. (Ibid.: 180 - 183)

The higher basic costs of PCN networks made it imperative for them to acquire a large
subscriber base very quickly, simply in order to survive. The cost problems associated
with DCS1800 could be solved by achieving sufficient economies of scale, but this
depended on a successful rapid diffusion of GSM -- a technology initially designed for
the ‘business’ market -- to the mass or ‘consumer’ market. The one technical advantage
of a DCS 1800 network, that it would have ”far greater capacity than its GSM
counterpart”, could only be realised if the operator could gain ”enough subscribers to
make it financially viable” (Garrard 1998: 183). From the user’s standpoint, however,
there were few technical distinctions to be made between the different systems developed
for the business and consumer markets. The basic problem that initially confronted PCN
could therefore formulated as follows: ”If DCS1800 operators cannot differentiate their
services from GSM using the technical characteristics of their networks, the enormous
up-front capital investments force them to find other, more commercial, ways to achieve
market success” (Garrard 1998: 184).

The ‘other ways’ to which PCN operators resorted in order to achieve rapid market
expansion involved lowering the price of subscriptions and network use. This depended
on, among other things, developing alternative tariff structures and re-negotiating
interconnection charges with the fixed networks, or PSTNs (Public Switched Telephone
Networks). The use of service providers by the networks enhanced their capability to
offer a wide variety of tariff structures, while lowering the costs of interconnection
depended primarily on the bargaining power of network operators. In addition, reaching
the mass market also required new distribution chains. These involved, for example,
consumer electronics stores rather than specialist shops as the main sales outlets,
improved procedures for registration, and far more extensive customer service support.
Finally, aggressive marketing, and the development of business organisations geared for
marketing rather than technical development, was also required to ‘grow’ the market.

Early competition along these lines in the UK, both among PCN operators and between
PCN and GSM operators, is extensively described by Garrard (1998: 184 - 217). He
concludes that PCN was successful in rapidly broadening the mobile telephone market to
a penetration rate of 12% by 1996, although at a considerable cost to network operators:
their profit margins remained narrow and highly vulnerable to consumer dissatisfaction
with the price and quality of service provision. One of the key lessons to be drawn from
the UK experience of a competitive market in mobile telephony is that a trade-off
between the extent of ‘coverage’ and the degree of service ‘quality’ was not acceptable to
consumers. ”The only way for a new entrant to succeed in the long term is to offer a
service that at least matches existing networks in every respect, as well as providing some
additional benefit” (ibid.: 217).

3.1.3. Paging Systems: ERMES

As mentioned above, in sub-section 2.3.2, paging systems continued to provide a
necessary complement to digital mobile telephones, due to their superior coverage
(Garrard 1998: 156). Paging systems ‘grew up’ alongside of mobile telephony, and
attempts to create a single European standard for paging systems date from 1970. At that
time, CEPT issued the Eurosignal standard – which, however, was adopted by very few
of its members (Paetsch 1993: 316).

The search for a pan-European paging standard continued without much real success
through the 1980s. It was only in the early 1990s that ETSI, having taken over this
project from CEPT, was able to persuade operators in about 15 countries to sign a
Memorandum of Understanding (MoU) committing them to the introduction of ERMES
(European Radio Messaging Standard) (Garrard 1998: 442). However, the first
commercial ERMES service was not initiated until very late in 1994 (ibid.: 444). By 1996
an alarmingly low rate of adoption – there were only 3 European ERMES networks in
operation at that time – led the European Commission to concede with some reluctance
that ERMES need not be recognised as the pan-European paging standard (ibid.: 445).

The failure of ERMES may have been largely due to the fact that long before this
standard was developed by ETSI, a, de facto world standard for paging systems had
already been established and widely adopted in Europe. This was the POCSAG standard,
whose acronym refers to the organisation that developed it -- the Post Office Coding
Standardisation Advisory Group. This group, representing a number of paging operators
and equipment manufacturers, developed the POCSAG system under the auspices of the
UK Post Office during the late 1970s (Paetsch 1993: 30 - 32). By the end of the 1980s,
variants of the POCSAG system were in wide use, not only in Europe but also in other
industrialised countries, except for Japan (ibid.). A significant drawback of the various
POCSAG systems in use was that they were incompatible with one another, and this
provided the rationale for attempts to create another, pan-European standard (ibid.: 302).

Eurosignal and later ERMES succeeded in defining such standards, but they were unable
to compete with the widely diffused POCSAG systems on cost (Garrard 1998: 433 -
434). ERMES, although it was technically superior to POCSAG in most respects, had
other problems as well. These included disruption caused to cable-tv networks by
ERMES transmitters, chronic shortages in the supply of pagers, the low demand for
paging systems in Europe, and the late development of the standard (ibid.:444 – 445).
When it finally emerged onto the world market in the mid-1990s, ERMES soon lost
ground to the US-based FLEX system developed by Motorola. FLEX was much more
widely adopted internationally than ERMES – due, in part, to the fact that it was
backwards-compatible with POCSAG, as well as being “smaller, simpler, and cheaper”
(ibid.: 445). Compared to only a handful of ERMES systems throughout the world in
1996, there were 21 Flex systems in the US and Canada, 12 in Latin America, and 39 in
Asia (including a large and rapidly growing national service that had recently been
established in China). In addition, some 53 manufacturers had licensed Flex technology

from Motorola. Recounting these statistics, Garrard (1998) concludes that “With the
advent of FLEX, prospects for ERMES seem bleak” (ibid.: 446).

3.1.4. Trunked Mobile Radio/Data:TETRA

Trunked mobile radio/data systems resulted from the evolution of earlier systems for
private mobile radio (PMR) and public access mobile radio (PAMR). A European
standard for both types of system eventually emerged in the form of TETRA (Trans-
European Trunked Radio). ETSI finalised the specifications for TETRA and then
published them for public inquiry in the fall of 1994, making TETRA a pan-European
standard whose development was roughly contemporary with the ‘second stage’ of GSM
(Garrard 1998: 424).

During the earlier years of the GSM system’s development – i.e., the late 1980s and early
1990s – some observers and analysts saw certain possibilities for inter-system
competition between cellular telephone systems and public access mobile radio systems
(PAMRs) (Lindmark and Granstrand 1995). Inter-system competition between digital and
analogue moblile telephony, and between these systems and alternatives such as PMR
and PAMR, was especially likely in the case of larger countries, such as Germany,
France and the United Kingdom, where multiple systems were in operation under
multiple licenses. In particular, PAMRs that had nation-wide coverage and access to the
PSTN network (public switched telephone network) were a “viable alternative to mobile
cellular networks”. This was, at least, the case for business users who needed to have
frequent contact with closed groups of other users but only occasional access to the PSTN
(Paetsch 1993: 319). It was also noted in this context that ” PAMRs that can be directly
connected to a customer’s PBX (private branch exchange) are capable of low-speed data
transmission, and furthermore have the advantage of a usage-neutral service-charge
structure, as opposed to the time-based service charge structure of cellular telephone
networks” (ibid.).

PAMR systems originated in earlier PMR systems, and their main advantages over the
latter were that they required less costly investments in infrastructure and provided more
extensive coverage. Moreover, PAMRs, because they were trunked systems, were also
more spectrum-efficient. The development of trunked mobile radio technology – and
especially publicly operated trunked mobile radio systems – was therefore promoted by a
number of European regulatory authorities as an effective response to growing problems
of spectrum-shortage in metropolitan areas (Hardiman 1990). During the late 1980s,
therefore, there was not only a steadily growing number of both private and non-private
PMR users in Europe but also an ongoing transition to public access trunked mobile radio
systems, or PAMRs. In this context, moreover, there arose a demand for standardisation.
“While proprietary standards were acceptable for traditional mobile systems, it is obvious
that PAMRs, operated by PTTs as well as independent organisations, require a well-
defined standard” (Paetsch 1993: 300). For these reasons – and also to co-ordinate the
development of PAMRs in Europe with that of GSM – ETSI set out, in the early 1990s,
to create a pan-European standard for digital-trunked PAMR systems that would be
completed by 1995 (ETSI Highlights 1989: 6).

TETRA was eventual result of this standardisation initiative. When TETRA emerged in
1995 and the first national licenses were awarded in 1996, however, it was clear that it
would pose no serious threat to GSM in terms of inter-system competition. As with
mobile data networks, such as the MOBITEX system discussed in subsection 2.3.2.,
TETRA and other PAMRs remained confined to a very limited market, comprised
primarily of business or public service users with special needs. The European
Commission, therefore, had little reason to regard PMR as a serious rival to GSM.

The EC noted, in its 1994 Green Paper on mobile and personal telecommunications, that
the diffusion of such systems faced two main barriers: “the general prohibition on
interconnection to the fixed network and customer resistance to sharing a network”
(Commission of the European Communities 1994: 110). Ultimately, proprietary PMRs
remained more attractive to business users. However, high costs and limited coverage
ensured that they would not become very widespread. PAMRs, although they solved
many of the technical problems that limited further growth of this type of system, would
not be widely adopted because they posed too many other drawbacks for “closed user
groups and corporate users” (ibid.). In 1995, there were less than 75,000 European
subscribers in total outside of the UK and Germany (Garrard 1998: 424). The largest
national market for PAMR in Europe was found in Germany, with a total of 140,000
subscribers in 1995, but this was “still dwarfed by cellular” (ibid.). In the second largest
market, the UK, there were still less than 100,000 subscribers in 1996 (ibid.: 423).

Despite its limited extent, the existence of a lucrative business market, together with its
co-ordinated development and compatibility with the GSM standard, ensured the further
evolution and eventual success of TETRA. The standard is now poised to become -- like
DECT, and for very similar reasons -- a platform for the development of ‘third-
generation’ mobile telecommunications technology. (For an account of the parallel case
of DECT, see subsection 3.1.1., above.) In 2001, the TETRA MoU adminsistration could
claim that it now represents “56 organisations, in 19 countries” and “the entire European
PMR industry, exclusive of MATRA and AEG” (TETRA MoU 2001: 1). According to
this same source, moreover, Tetra has now reached “a stage of completion from ETSI and
a 100% acceptance from all the European Administrations” (ibid.). In 2001, ETSI
reported that TETRA standardisation had reached “a mature state” (ETSI 2001b: 3). It
further declared that planned improvements, in areas such as high-speed transmission of
packet data, intercommunication with other systems, and further enhancements would
“provide improved interworking and roaming between TETRA and public mobile
networks such as GSM, GPRS and UMTS” (ibid.).

3.2. Rival Standards

This section will address only ‘rival standards’ representing competing technologies of
the same type as GSM – that is, second-generation (digital) systems for cellular mobile
telecommunications. As distinct from ‘inter-system’ competition, wherein different
technological systems that are only partially capable of performing the same functions

contend with one another for ‘market share’, the type of competition considered here
occurs between technological systems of that are of the same type.3 Such systems may
have different technical specifications, but they do not differ fundamentally with respect
to their main functions (Lindmark and Granstrand 1995: 380 - 381). The competition
occurs between alternative technical standards proposed for a given type of system, and
in highly regulated markets such as telecommunications, “the adoption of a system [i.e.,
standard] on the market of end-users is essentially a two-stage adoption, since the system
first has to be adopted by the regulatory body, handing out concessions” (ibid.: 318).
Thus, the acceptance of one or another standard by regulators is the first battle to be won.

There were three main rival standards for digital cellular telephony in the second
generation of mobile telecommunications. In addition to the European GSM standard
there were also the US-based D-AMPS (digital AMPS) and CDMA standards and the
Japanese PDC (Pacific Digital Cellular) standard. The discussion below will deal briefly
with each of these rival systems. It will not, however, provide any detailed account of any
complementary standards such as those already discussed in section 3.1.

3.2.1. D-AMPS

As in the case of European authorities and the GSM standard, the decision of the US
Cellular Telephone Industry Association (CTIA) to develop a digital standard was
motivated primarily by capacity problems with pre-existing analog systems. There was a
brief initial debate on what technology the standard should be based on. The main
alternatives were the FDMA (Frequency Division Multiple Access) technology backed
by AT&T and Motorola and the TDMA (Time Division Multiple Access) technology
supported by Ericsson and Northern Telecom, among others. The question was resolved
by 1990 in favour of the TDMA technology (Lindmark and Granstrand 1995: 390). This
formed the basis for the CTIA’s publication that year of the IS-54 standard (ibid.). IS-54
eventually became more widely known as the D-AMPS (digital AMPS) standard.

In contrast to the European decision regarding GSM, the CTIA opted for compatibility
with first-generation systems, based on “two seamless transitions – from analog to a six
time-slot digital system followed by a shift to a twelve time-slot system” (Paetsch 1993:
390). This decision made possible an incremental shift from first to second generation
technologies, based on the gradual conversion by carriers of the radio frequencies
previously allocated to AMPs systems. “This approach, however, necessitate[d] not only
a dual cellular infrastructure but also a dual-mode cellular phone for the time of transition
…” (ibid.). This terminal equipment, in addition to being bulkier and heavier than either

  Hence, the discussion here does not consider higher levels of system competition, such as that between
’wired’ and ’wireless’ telecommunications technologies. This is a topic more appropriately dealt with in
relation to ’third generation’ mobile telecommunications, given that the displacement of ’wired’ by
’wireless’ telecommunications not become a clearly defined trend prior to the advent of third generation
technologies (Lindmark and Granstrand 1995: 398 - 399). Similarly, the discussion here does not take up
competition between first and second generation technologies for mobile telecommunications. It has long
been established that the first generation – which is to be dealt with in a separate paper – was superceded
by the second (Lindmark and Granstrand 1995: 397). Reasons for this have been discussed in sections 1
and 2.

the analog equipment that it replaced or the all-digital handsets that would eventually
come to replace it in turn, was also more costly (ibid.). Moreover, the system as a whole
required considerable time for development, and did not become commercially viable
until 1993, even after which adoption rates remained slow (Lindmark 1995: 390). In
these respects, D-AMPS was a laggard compared to GSM.

Even before the CTIA decided on the IS-54 standard in 1990, regulatory decisions and
the development of competing standards had created the basis for ongoing uncertainty
about the emergence of any single ‘second generation’ standard in the US. The FCC
(Federal Communications Commission), as the relevant regulatory authority in the US
had already ruled, in 1988, that there would be no national digital standard defined for the
US. It only insisted that existing requirements for national compatibility would continue
to apply to any of the new cellular systems that carriers were free to adopt (Garrard 1998:
315). This requirement provided the underlying rationale for CTIA’s decision to opt for a
‘backwards-compatible’ standard in the form of IS-54, despite the drawbacks that it
incurrred of costliness and delay. In addition to these hindrances, D-AMPS also had to
face competition from other domestic standards, another outcome of the FCC ruling.

As discussed below, in section 3.2.2., CDMA, the main domestic competitor to D-AMPS
in the US, emerged several years later than D-Amps. Nevertheless, many of the network
operators that had originally opted for the D-AMPs standard subsequently changed their
allegiance to CDMA. In 1996, therefore, there were only three major network operators
that continued to D-AMPS, whereas eight operators had chosen to use the CDMA
standard instead (Garrard 1998: 324). Despite this split, D-AMPS still retained a large
share of the US market, due to the size of the operators (including AT&T Network
systems) that continued to use it. Thus, “TDMA could potentially serve areas with a
population of 204 million; major operators choosing CDMA served 256 million” (ibid.).

This partition of the market meant, in effect, that there would be only limited competition
between the two standards. In some areas of the US, operators offering both standards
were in competition. But in other areas, many end-users would never have a real choice
between the two digital standards because only one of the options would be made
available to them.

There were additional problems of a more technical character that also ensued from the
manner in which the two standards had evolved. Neither one was directly compatible
with the other, although both were ‘backwards compatible’ with pre-existing analogue
standards in the US. In 1998, therefore, it appeared that “roaming between one network
with CDMA and another with TDMA will only be possible using analog channels”
(Garrard 1998: 325). It also appeared that that “since the FCCC requirement for nation-
wide compatibility still remains in force, a complete migration to digital services seems
impossible from a regulatory point of view, even if operators decided that they wanted to
pursue that option” (ibid.: 325 – 326).

The capabilities of specific systems that were developed were not only compromised with
respect to ‘roaming’. There were other flaws as well. In the case of D-AMPs, the

‘backwards compatibility’ of the standard with the pre-existing analogue standard,
AMPS, meant that D-AMPS was not an entirely new system and therefore did not require
extensive infrastructural development in the form of “installed base” (Funk 1998: 434).
Essentially a digital extension of its analogue predecessor, AMPS, D-AMPS was only
installed in large cities whose existing AMPS networks required increases in capacity.
These circumstances greatly limited the technological development potential of D-
AMPS. “This meant that although D-AMPS would be cheaper for US carriers to
implement than a completely new system, it would not be as technologically
sophisticated as a completely new digital system such as GSM “ (ibid.).

For a number of reasons, then, neither D-AMPS nor any other US-based standard for
second generation mobile telecommunications succeeded in dominating the domestic
market in the US. This outcome was, moreover, linked to technical shortcomings that had
repercussions for the performance of the US standards internationally. Compared to the
rapid and extensive international diffusion of the GSM standard (see section 1.3), the US
standards did not fare well. In the case of D-AMPS, the main market development
outside of the US occurred in Latin America and some parts of Asia:

           All the Latin American countries adopted AMPS/D-AMPS as their national
           standard, often insisting that new operators introduced digital technology
           from the start of service instead of waiting until they needed additional
           capacity. A number of Asian countries licensed D-AMPS services alongside
           GSM, following their policy of allowing technologies to compete, as well as
           competitors. (Garrard 1998: 327).

3.2.2. CDMA

The emergence of domestic standards for ‘second generation’ mobile telephony that
would compete with D-AMPS was partly linked to the FCC’s 1988 decision. However, it
was also due to the character of the Telecommunications Industry Association (TIA), the
organisation finally responsible for the development of telecommunications standards. As
already noted, the CTIA – an organisation which represented not only manufacturers of
mobile telecommunications equipment but also ‘carriers’ or network operators and other
users – played an important role in defining mobile telecommunications standards in the
US. Its proposals, however, also had to be ratified by the TIA (Paetsch 1993: 143).

The TIA was essentially “a manufacturer’s trade association” that possessed a higher
degree of authority than CTIA, due to the fact that it had accreditation from the American
National Standards Institute (Garrard 1998: 315). In composition, TIA was quite unlike
its European counterpart ETSI, which was dominated primarily by ‘carriers’ (i.e., the
PTOs and PTTs). TIA was much less influenced by carriers’ demands, which it only
needed to take into consideration when developing what were essentially ‘voluntary’
standards (Paetsch 1993: 143). This difference in organisational structure resulted in a
much more pluralistic approach to standards development in the US case, based on the
need to accommodate competing interests among manufacturers.

1n 1990, at the same time that the CTIA decided on IS-54, Motorola proposed a narrow-
band analogue technology called N-AMPS that was capable of overcoming the capacity
problems of first generation cellular telephony (Garrard 1998: 318). The N-AMPS
proposal failed, but another, more serious contender had already emerged.

In 1989, even before N-AMPs was announced, the US telecommunications equipment
manufacturer, Qualcom, presented the CTIA with an alternative to the TDMA technology
that was the basis for IS-54. This alternative was CDMA (Code Division Multiple
Access), a “wideband” technology which Qualcomm continued to develop as a
proprietary standard, even after the CTIA had decided in favour of TDMA as the
‘industry standard’. Qualcomm was joined in this effort by a number of major
organisations such as Motorola, AT&T, Nynex, Ameritech, and PacTel (ibid.: 317 –
318). Among the major equipment manufacturers, the only firm not to join the coalition
was Ericsson, “which continued to state publicly that it did not believe that CDMA
offered any advantages over GSM and that it did not intend to manufacture any new
products for the new technology” (ibid.: 319).

Apart from this exception, Qualcomm had been able to mobilise strong support for its
CDMA proposal, especially among equipment manufacturers. Hence the CTIA and TIA
were faced with a serious quandary, which the CTIA eventually resolved by affirming its
original decision in favour of IS-54 but also resolving to continue work on a “wideband
standard”. This compromise led, eventually to CTIA’s adoption in 1994 of CDMA as
standard IS-95, which was “to become the second industry cellular standard in the USA”
(ibid.: 319).

CDMA thus emerged considerably later than D-AMPS as an ‘industry standard’, and was
therefore not initially a direct competitor to D-AMPS. In the longer term, though, it did
become a competitor, and the more lengthy period of time required for its development
enabled CDMA to gain the backing of an extensive consortium of telecom carriers or
operators and equipment manufacturers. In addition to this, CDMA possessed a number
of features that made it technically superior to D-AMPS. For example, CDMA was
capable of capacity gains on the order of “a twenty-fold increase over the AMPS
system” (Paetsch 1993: 179). These aspects of CDMA’s development provide some
explanation of why so many major network operators switched rather quickly to the new
US digital standard when it finally emerged. (See sub-section 3.2.1., above.)

In spite of the advantages gained from its delayed launch, CDMA still suffered from
problems of technological “immaturity” and the “limited availability” of terminal
equipment when it was finally introduced by those operators who had chosen to back this
standard (Garrard 1998: 325). Like D-AMPS, CDMA achieved “less growth in installed
base … than if such a standard were installed throughout the US” (Funk 1998: 435).
CDMA was also implemented differently by each operator. Thus, like D-AMPS, its
potential impact on the US market as a whole was seriously limited by an initial
partitioning of the market based on operators’ choice of one or the other standard (see
sub-section 3.2.1.). This, in turn, limited the standard’s technological development. And,

as a consequence, the CDMA standard, like D-AMPS, met with only limited success

Outside of the US, the most notable adoptions of CDMA occurred in Asia. The world’s
first CDMA system was launched in Hong Kong, in 1995, and had gained some 10,000
subscribers by the end of that year (Garrard 1998: 325). Even earlier, South Korea had,
for reasons of industrial policy, adopted CDMA as its national standard in 1993. This
decision which eventually led to the launch of two CDMA service in 1996, and several
more thereafter, as well as the licensing of CDMA technology to a number of Korean
manufacturers of telecommunications equipment (ibid.). As of 1998, CDMA also had a
number of other good prospects in the Asia-Pacific region, but these would involve
serious competition with other standards – most notably GSM, which had already been
implemented in these countries. Thus, while it seemed “very likely that a number of
countries … may license additional operators for CDMA”, it also had to be noted that
“many have already licensed almost every other technology that became available” (ibid.:

3.2.3. PDC

The Japanese contribution to digital cellular telephony in the second generation of mobile
telecommunications was the PDC (Pacific Digital Cellular) standard, which never
diffused widely out of its home country, Japan. This insularity has been variously
attributed to the standard’s lack of “openness” (Funk 1998: 434) and Japanese
telecommunications equipment manufacturers’ “lack of a track record in international
markets” (Garrard 1998: 326). In any case, PDC remained essentially a domestic

The development of PDC began with a 1989 decision by Japan’s Ministry of Posts and
Telecommunications (MPT) to create a high-capacity digital standard for cellular
telephony (Garrard 1998: 326). The technical requirements were initially defined by an
internal Ministry working group, the Japanese Digital Cellular Radio System Committee,
and the standard itself was developed by subcommittees of the Standards Committee of
the Japanese Research and Development Centre for Radio Systems (RCR) (ibid.: 361).
The standard that was eventually developed by 1990 had many basic features that were
highly similar to D-AMPS. The main differences were due to spectrum allocation policy
(ibid.). As a result of these differences, though, PDC would be incompatible with any
other systems. “Although Japan’s MPT did require … the same TDMA technology that
GSM and D-AMPS are based on, the frequencies that it allocated for digital cellular
systems were not compatible with these systems” (Funk 1998: 435).

The PDC system was first launched in Tokyo in 1993 by NTT DoCoMo, which was once
the mobile subsidiary of NTT, the national telecommunications operator, but had since
become a separate company (Garrard 1998: 326). This launch was soon followed in
1994, the same year that Japan’s market for terminal equipment was liberalised, by the
introduction in other densely populated areas of two further large digital cellular
networks. Each of these networks was a joint venture of one of two newly formed

consortia – Tu Ka and Digital Phone – whose composition included Japanese subsidiaries
of international telecommunications equipment manufacturers (e.g., Motorola) and
operators (e.g., Cable and Wireless) in addition to their domestic counterparts (Garrard
1998: 362). These new networks were soon joined by others, with the result that NTT’s
network quickly found itself facing a number of competitors.

The fact that PDC was launched under competitive market conditions contributed to
slower growth than expected for NTT, but a fairly rapid adoption of PDC networks
within Japan. In 1995, competition had become intense and the growth of the market for
mobile telecommunications had become rapid. In that year, the price of terminals for the
new PDC standard had fallen below $500 US, and in the last quarter all operators cut
their monthly fees and registration charges. As a consequence, “by the end of the year
there were nearly nine million subscribers, with 60% of them connected to digital
networks” (Garrard 1998: 363).

As mentioned earlier, however, PDC never flourished outside of Japan. In addition to the
reasons mentioned above (lack of ‘openness’ and Japan’s weak presence on international
markets for telecommunications equipment), the lack of success abroad was at least
partly due to NTT DoCoMo’s continuing domination of the standardisation process that
created PDC. One observer has noted that “since NTT DoCoMo was not allowed to
operate services overseas, it had little interest in creating a world-wide standard” (Funk
1998: 435). Consequently, NTT DoCoMo “used its control of the standard-setting
process to create a relatively closed, unique standard that is still only used in Japan”

3.3. Summary

This section has dealt with both complementary and rival standards related to GSM. It
has pointed out that both kinds of standards have been competitors to GSM, but only to a
limited extent.

Section 3.1, on complementary systems and standards, discussed other types of
technological systems that might originally have constituted ‘competitors’ to systems like
GSM. These, in some cases at least, could have provided partial substitutes for GSM and
other second-generation (digital) systems for cellular mobile telecommunications, in
areas such as data-transmission, telephony or voice transmission, and short messaging (or
paging). In most of these cases, though, the only partial potential for substitution was
eventually translated into the development of complementarities between systems. In
several such cases within Europe, ETSI was able to direct these systemic
complementarities towards the evolution of standards compatible with GSM.

ETSI thus created a whole ‘family’ of compatible technological standards. DECT, DCS
1800, and TETRA were not only complementary to, but also compatible with, GSM.
Accordingly, they both benefited from and helped to advance and consolidate GSM’s
rapid domination of the European market for mobile telecommunications. ETSI’s most

notable failure, ERMES, was a complementary system, but could gain no special
advantage from GSM compatibility. Moreover, it fared poorly in relation to other systems
of the same type that had superior technical or market features. None of these paging
systems, however, was a viable substitute for mobile telephony. Therefore, none
detracted from the technological and market development of systems such as GSM.

Section 3.2. dealt with rival standards for systems of the same type as GSM. It showed
that there was never a full competition among these standards, since the initial decisions
made by regulatory authorities effectively partitioned the international market. Japan was
perhaps the most extreme case. There, close relations between the standard-setting
authority and the national telecommunications operator in charge of developing the new
technology led to the latter’s effective domination of the standardisation process and the
creation of a closed national standard. Conversely, the market remained closed to other

Even in the case of the US, which had a competitive policy regarding standards, the entry
of foreign standards was minimal, and the domestic market was effectively partitioned
only between two rival domestic standards. This outcome ensued not only from
regulatory decisions but also from the decentralised and ‘voluntary’ nature of standard-
setting and standards development in the US context. D-AMPS – and, for that matter,
any other second generation standard – never managed to dominate the US market
completely. Similarly, and partly as a consequence of this domestic rivalry, neither D-
AMPS nor CDMA ever emerged as a dominant force in international competition. This
pattern is clearly illustrated in Table 3.1, below, which describes the distribution of
market shares for the major second generation standards in mobile telecommunications at
the end of the 1990s:

Table 3.1: International Competition among Major ‘2nd Generation’ and Other
Mobile Telecommunications Standards: Regions and Market Shares

     Region    Europe          North          Latin           Asia               Africa
Standard                       America        America
GSM            89%             4%             1%              35%                88%
D-AMPS                         27%            39%             3%
CDMA                           9%             9%              14%
Analogue       11%             60%            51%             48% (includes      12%
                                                              PDC & PHS –
                                                              Digital Cellular

(Source: ITU World Communications Database)

As Table 3.1 reveals, the D-AMPS standard captured a large share of international
markets only in Latin America, and the CDMA standard made somewhat more limited

headway in Asia, where D-AMPS had little success but GSM had far more. (The reasons
for these developments were discussed at some length in sub-sections 3.2.1 and 3.2.2.)
Table 3.1 also indicates the ‘regional’ character of the market for mobile
telecommunications. What is perhaps most notable about the limited successes of US
standards abroad is that in both cases the US-based standards fared better internationally
than domestically. This was due not only to a ‘divided market’ but also to the weaker
migration in North America than elsewhere from ‘first generation’ to ‘second generation’
standards. (For explanation, see again the discussion in sub-sections 3.2.1 and 3.2.2.) By
contrast, the GSM standard, based in Europe, the region of most extensive migration, not
only dominated the domestic market but also the African market. In addition, GSM
captured 35% of the Asian market and even made some inroads on the North American
Market (See section 1.3.)

4. Public Sector Actors

Compared to the development and implementation of the NMT standard(s), the
emergence of the GSM standard was marked by the involvement of a far greater number
of actors and a much greater complexity in the relations among them. The playing field
was no longer one region within Europe but, instead, Europe as a region of the world.
For public sector actors -- in particular, the PTOs or PTTs -- the 1980s and 1990s were
years of transition. Until the 1980s, the PTOs and PTTs exercised extensive monopoly
powers, often combining the roles of regulation, network operation and service provision.
By the mid-1990s, their roles had become much more specialised. Increasingly, the
functions of PTOs and PTTs were confined to network operation, and even in this area
their monopolies were progressively eroded. New organisations, such as separate
regulatory authorities, had been created to take over many of their former functions, and
they faced domestic competition from new service providers.

Similarly, private sector organisations were increasingly exposed to higher levels of
international competition. In addition, the convergence of formerly separate technologies
in the ‘digitisation’ of mobile telecommunications meant that incumbent firms in the
telecommunications sector now had to confront new entrants whose competence had
originated in other sectors. Strategic alliances (among both public and private sector
actors) became prevalent. Moreover, firms also began to rely increasingly on public-
sector research. Rather than collaborating exclusively with PTOs or PTTs,
telecommunications equipment manufacturing firms began joint R&D with universities
and research institutes. All of these developments had far-reaching implications for the
institutional framework and knowledge-base of second-generation mobile telephony, as
indicated in the following remarks:

            The technological base has broadened to include innovations from outside the
            traditional telecommunications sector -- chiefly from computer and software
            firms -- that have become part of the fabric of the contemporary network.
            Vertical ties between PTOs and selected equipment manufacturers have also
            loosened. Manufacturers now seek new international markets to
            counterbalance the increased R&D expenditures that new digital product lines
            have incurred. Likewise, even if they remain intent upon preserving their
            monopoly positions, many PTOs still see economic advantages in expanding
            their range of possible suppliers. Factors such as these have virtually
            required that the standardisation process be opened up to some extent.
            (Hawkins 1995: 29 - 30)

4.1. European National PTOs or PTTs and SDOs

As with the first-generation standard, NMT, public sector actors were very important to
the development of the second-generation standard, GSM. However, there were
discontinuities, as well as continuities, with respect to the involvement of public sector

actors. One of the main continuities was the leading role played by national PTOs or
PTTs in initiating and carrying out the project of defining a new standard. But there were
also discontinuities in the part played by these organisations. For one thing, there were
many more PTOs and PTTs involved in the development of GSM than in the case of
NMT. For another, they did not co-operate solely on an ad hoc consortium basis, but
rather within a formal organisational framework provided by two European SDOs.

The first of the SDOs associated with GSM was CEPT (the European Conference of
Posts and Telecommunications), which initiated GSM. The second was ETSI (the
European Telecommunications Standards Institute), which saw the GSM standard
through to its completion. The transfer of responsibility for GSM and other
telecommunications standards from CEPT to ETSI was also marked by a shift from a
‘closed’ to a more ‘open’ approach to standards development. The more inclusive
approach of ETSI, which welcomed equipment suppliers and public research
organisations, among others, as members, reflected a growing realisation that PTOs and
PTTs could no longer claim or exercise a monopoly of knowledge and expertise in

4.1.1. The Changing Role of PTOs/PTTs in Europe

As in the case of NMT, the PTOs or PTTs were leaders in developing the new standard.
NMT, however, had been the initiative of only a few Nordic PTOs or PTTs and allied
organisations. Many more PTOs and PTTs were involved in developing GSM, since this
was an initiative of CEPT (the European Conference of Posts and Telecommunications),
an organisation which represented some 25 European PTOs or PTTs.

There was little change in the character of these organisations, at least during the
development period for ‘phase 1’ of the GSM standard, which was introduced in 1992.
The majority of the members of CEPT were state-owned monopolies that combined
several roles (Hawkins 1995: 29). Most were both network operators and service
providers. In addition, many also combined these roles with that of regulating the
telecommunications markets in their respective countries. Thus, the PTOs or PTTs
exercised a very high degree of ‘market power’, although this state of affairs was
conventionally justified by the understanding that telecommunications formed a ‘natural
monopoly’ (Muller 1992).

As monopolists in ‘closed’ national markets, the PTOs or PTTs also had special
responsibilities for the development of new technology and the improvement of market
conditions in the telecommunication sector. The first part of this mandate, especially,
often involved ”strong co-operation between the incumbent operators and the national
telecommunications industry; in some cases this was almost tantamount to a kind of
vertical integration” (Pisjak 1994: 289).

 By 1992, however, when GSM was first introduced, it had become apparent that the
‘rules of the game’ would change dramatically for PTOs or PTTs within the next few
years. In the European Union, the publication in 1987 of the EC Green Paper on

Telecommunications (Commission of the European Communities 1987) had profound
implications for these organisations. Among other things, it spelled the end of a policy
regime in which domestic telecommunications standards were ”defined internally by
monopoly PTOs, sometimes in collaboration with preferred suppliers” (Hawkins 1995:
29). As mentioned above and explained below, the EU’s creation in 1988 of ETSI, which
took over the responsibility for standard-setting from CEPT, marked a decisive shift from
a ‘closed’ to an ‘open’ approach to developing standards.

The EC Green Paper also initiated a period of extensive liberalisation of
telecommunications markets in the EU, paralleling similar developments across the
OECD world. The reforms that were initiated soon afterwards included the opening of
formerly closed national markets to international competition and the end of monopoly
privileges for the PTOs or PTTs. New service providers were also permitted to emerge
within national markets. Other reforms included the creation of separate bodies to take
over the function, formerly carried out by many PTOs or PTTs, of market regulation.
Liberalisation was a complicated and therefore gradual process, beginning earlier in some
countries and later in others. Nevertheless, by the end of the 1990s the liberalisation of
telecommunications networks in Europe was at an advanced stage, even if it was not yet
fully completed (OECD 1999a: 11 - 12).

4.1.2. The Swedish Case: Televerket

As noted above, the national PTOs or PTTs traditionally had a mandate to improve
telecommunications technology. They also had strong historical linkages with national
telecommunications equipment manufacturing firms, who tended to be their primary
‘partners’ in collaborative R&D efforts. A prime example was the Swedish PTO,
Televerket (later, Telia), and its close collaboration with the Swedish telecommunications
equipment manufacturer L. M. Ericsson (later, Ericsson). Televerket was charged with a
mission to develop new telecommunications technology, and to support national
equipment suppliers in doing so (Karlsson 1998: 30 - 32, Chap. 3). For its part, L. M.
Ericsson’s success in developing major innovations in telecommunications equipment
had, at least until the development of NMT, depended strongly on its relations with
Televerket. In addition to constituting a vital ‘test market’ for Ericsson, the Swedish
PTO had ”provided crucial resources for the innovation activities through technical and
financial risk-sharing and technical collaboration activities” (Fridlund 1998: section 6).
The close collaborative relationship between Televerket and Ericsson was continued in
the initial development of GSM (McKelvey, Texier, and Alm 1998: section 9.1).

The GSM consortium as a whole was comprised of national PTOs or PTTs and the
equipment manufacturers with whom they chose to collaborate. Beneath the broader
‘umbrella’ of the consortium, a number of different alliances or coalitions were formed
between these two types of actors. Thus, five such alliances submitted eight prototype
technologies to CEPT’s initial ‘technology competition’ for GSM, which was conducted
in order to establish a technical basis for further development of the first set of
recommendations for the standard (Garrard 1998: 129). The five coalitions were the

1 - Televerket, the Finnish, Danish and Norwegian PTTs, Ericsson; Nokia, and Elab (a
Norwegian firm)
2 - ART, SAT, SEL, AEG, Italtel
3 - Philips and TRT
4 - LCT; and
5 - ANT, Bosch and Telettra
(Lindmark 1995: 111)

It can be seen from this list that PTOs or PTTs were heavily involved in the development
of the ‘Nordic’ alternative that was considered for GSM standard. However, their
counterparts in other countries also played an important role with respect to competing
alternatives. Although there were five alliances in all, these could be further grouped into
two main ‘camps’: “a Franco-German group and a narrow-band camp championed by
actors in the Nordic region” (Glimstedt 2000: 9). The Franco-German group, which had
developed a series of ‘wideband’ or ‘broadband’ solutions based on large investments in
wideband-TDMA, during the 1980s, included France Telecom and its German
counterpart, DPT. It also had the backing of Alcatel and Siemens, the major
telecommunications equipment manufacturers in France and Germany, respectively.
(Ruotto 1998 257 - 258) National industrial interests were, moreover, overtly promoted
via the French-German ‘wideband’ proposals [Cattano, 1994 #918].

Sweden’s Televerket played an important ‘lead role’ in the opposing camp. The ‘Nordic’
coalition, of which both Televerket and Ericsson were members, submitted four separate
proposals for a ‘narrow-band’ GSM prototype technology. (All other proposals were for
‘wideband’ solutions.) Within this alliance, ”Televerket was the only PTT offering an
alternative; all the rest were firms” (McKelvey, Texier, and Alm 1998: section 9.1).
Televerket’s alternative, moreover, was the one selected. This decision conferred an
early technological lead on Ericsson, Nokia and other Nordic equipment manufacturers in
the commercial development of GSM technology. Conversely, ”If a broadband solution
would have been chosen as the new European digital standard, the Nordic actors would
have been left way behind in the technological development” (McKelvey, Texier, and
Alm 1998: section 9.1).

One might infer from this evidence that Televerket was not only a prominent sponsor but
also a main author of the technology that became the prototype for GSM. As emphasised
by Glimstedt (2000: 8), however, there is also much evidence to indicate that, in fact,
Televerket acted less as a ‘spearhead’ and more as a ‘figurehead’ for the allied private
firms, who had already begun to take the lead in technological development.

The achievements of engineers at Televerket and Ericsson, although notable, were based
on already well established basic technologies to which a number of non-Swedish firms –
primarily, Motorola, AT&T, Bull and Phillips – held the intellectual property rights
(Bekkers, Duysters, and Verspagen 2000). Moreover, the US-based firm, Motorola, held
many of the most important of these patents, which it licensed selectively to the main
Nordic equipment manufacturers, Nokia and Ericsson. In the case of Ericsson, Motorola

made cross-licensing agreements in return for access to Ericsson’s previously developed
expertise in digital switching. (Ibid.) Motorola’s strategy, vis-à-vis GSM, of using
licensing agreements as a means of promoting its interests and generating revenue-
streams based on its R&D investments, is well-documented (Iversen 2000).

Arguably, then, Motorola and the firms with which it made cross-licensing agreements
were probably the most important sources of technological advantage for the successful
‘Nordic’ GSM prototype. Conversely, evidence to this effect detracts from the
technological leadership that might be attributed to Televerket and other Nordic PTOs or

In addition to helping develop the GSM standard, and especially the prototype
technology on which it was based, Televerket also played an important role as a ‘user’ of
GSM – i.e., as an operator of GSM networks. However, this role coincided with the
privatisation of Televerket and its transformation into the public corporation, Telia. This
second aspect of Televerket/Telia’s involvement with GSM is therefore dealt with in a
later section (5) on ‘private sector actors’. (See, in particular, sub-section 5.2.3.)

4.2. European SDOs (CEPT and ETSI)

As previously noted, two European SDOs -- CEPT and ETSI -- played prominent roles in
the development of the GSM standard. CEPT (the European Conference of Posts and
Telecommunications) initiated work towards the definition of the standard. ETSI (the
European Telecommunications Standards Institute) took over CEPT’s mandate for the
development of telecommunications standards and completed the implementation of
GSM. The MoU group, which was originally formed within CEPT to administer the
Memorandum of Understanding on GSM and later migrated to a new home in ETSI is
discussed below under a separate sub-heading.

4.2.1. CEPT (Conférence Européenne des Administrations des Postes et

CEPT, formerly the primary organisation responsible for the development of
telecommunications standards in Europe, has been described as an ‘exclusive club’. Its
membership during the early 1980s, at the time when the GSM standard was first
proposed, was comprised of the PTOs and PTTs of some 25 European countries. Many
of these organisations were state-owned, and most were monopolists. In addition, many
members of CEPTalso acted as regulators of telecommunications within their ewspective
countries. These aspects of the organisation created a certain degree of conflict and
tension with respect to its mandate -- especially after the European Commission
requested CEPT, in 1975, to ensure principles such as ‘market co-ordination’ in its work
on standards (Drake 1994; Hawkins 1992). For one thing, CEPT had no power to enforce
the standards that it set (Wallenstein 1990). Also, and perhaps more importantly, “the
CEPT members were reluctant to meet these requests because such harmonisation

attempts conflicted with their national sovereign powers” (Bekkers and Liotard 1999:

The European Community’s ‘new approach’ to technical harmonisation and standards,
implemented during the 1980s, could have ameliorated these problems. It made possible
not only a stronger regulatory role for CEPT at the international level but also a break
with past methods of ‘consensus’ based decision -making, which had been identified as
an impediment to effective standard-setting (Schreiber 1991: 99). Nevertheless, the
members of CEPT continued to agree upon standards and measures for technical and
administrative harmonisation through processes of internal negotiation. In this manner,
some members -- particularly those from the Netherlands and the Nordic countries --
persuaded CEPT to establish the Groupe Spécial Mobile (or GSM) in 1982 for the
purpose of defining a new, pan-European mobile telephone standard (Garrard 1998: 126).

The Groupe Spécial Mobile later migrated, in 1989, to a new organisational base in ETSI,
where it was renamed the Special Mobile Group (or SMG). However, it carried out most
of the initial work involved in defining the GSM standard under the auspices of CEPT.
These activities began with the drafting of recommendations for the new standard in the
early 80s. They also included the testing, during the mid-80s, of prototypes submitted in
order to develop a more complete specification. In addition, they involved the
administration of the standard, once it had been defined, in the late 80s. Under the
direction of the European Commission, the GSM formulated in 1987 a Memorandum of
Understanding (MoU), initially signed by the PTOs or PTTs of some 13 European
countries. The MoU committed the signatories to a number of regulatory conventions
and a schedule for the implementation of the GSM standard. (Garrard 1998: 129 - 131).

The Groupe Spécial Mobile sought to include equipment manufacturers in its activities at
a very early stage in the process of standard development, although this was done only
‘by invitation’. Thus, the prototype systems that were submitted in 1986 to meet the
initial draft recommendations for the GSM standard were produced by different alliances
of PTOs or PTTs and collaborating telecommunications equipment manufacturers
(Lindmark 1995: 111). (See sub-section 4.1.2.) The GSM’s inclusion of manufacturers
marked a significant departure from the past practice of CEPT, whose membership
structure ”precluded direct participation by non-PTO interests” (Hawkins 1995: 29). It
also prefigured the ‘open’ approach to standards development that was to become
common among SDOs by the end of the 1980s. Nevertheless, CEPT remained an
organisation dominated by PTOs or PTTs, in significant contrast to the TIA, its
counterpart in the US, which was essentially an organisation of equipment manufacturers.
(See sub-section 3.2.2.)

Given the autonomous character of CEPT and the considerable power of its member
organisations, it is remarkable that the European Commission, which has been aptly
described as “lacking voice within CEPT” (Glimstedt 2000: 4), was able to intervene
effectively in its internal decisions regarding the development of the GSM standard. Yet
the EC not only managed to wrest authority over standards development from CEPT and
transfer it to the newly created ETSI in 1988. Even earlier, in 1987, it had been able to

direct CEPT’s formulation of the GSM Memorandum of Understanding (MoU).
Arguably, the EC was also able to influence CEPT’s choice, in the same year, of the
technological basis for the GSM standard. (See sub-section 4.1.2.) In that competition,
the ‘wideband’ coalition, including some of the most powerful members of CEPT, found
itself very isolated. It lost by a very broad margin, with 13 of CEPT’s 15 members voting
instead for the Nordic ‘narrowband’ alternative (Ruotto 1998).

The main leverage that the European Commission held in relation to CEPT was of a
regulatory character – or, more specifically, its determination to liberalise Europe’s
internal telecommunications markets after the model of recent deregulation in the US
(Sandholtz 1993). As noted in sub-section 4.1.1., the publication in 1987 of the EC
Green Paper on Telecommunications (Commission of the European Communities 1987)
made these intentions explicit. In addition to liberalisation, it heralded an end to using
standards development in telecommunications as a vehicle for purely ‘national’ industrial
policy interests within organisations such as CEPT. These principles, moreover, were
already in force, as part of the EC’s ‘single market’ policy (Bach 2000). They had been
previously established by the Commission in 1986, through Council Decision 22,
requiring the use of European Standards in the public procurement of information and
telecommunications technologies (ibid.).

Given the powerful mechanisms that could eventually be brought to bear on CEPT and its
member organisations by the European Commission, CEPT had little choice but to
accede to the EC’s initiatives regarding standard setting in general and its specific
suggestions regarding the further development of the GSM standard. Similarly, CEPT
had little interest in making any decision regarding GSM that would obviously privilege
its most powerful members and the ‘national champions’ associated with them.

4.2.2. ETSI (European Telecommunications Standards Institute)

The EU established ETSI in 1987-88 as a standards-development organisation (or SDO).
As noted earlier, the creation of ETSI was initially suggested by the European
Commission, which had become increasingly concerned about Europe’s competitive
position in telecommunications and the threat posed by US multinationals entering
Europe via new joint ventures with European companies (Dang-Nguyen, Schneider, and
Werle 1993). The EC had expressed concern in its 1987 Green Paper on
Telecommunications that the EU was not making a sufficiently strong and concerted
effort in the development of telecommunications standards to meet the demands of the
internal market (Commission of the European Communities 1987). GSM was, of course,
a notable exception. For this and other reasons, such as the fact that it was issued at a
relatively late stage in the development of GSM, the 1987 Green Paper ”set mobile
telecommunications to one side for future consideration” (Commission of the European
Communities 1994: section 1).

ETSI took over from CEPT the work of formalising and implementing GSM. By 1988,
when ETSI came into being, the recommendations for ‘phase 1’ of the GSM standard
were nearly complete. Nevertheless, work on the specifications in various Technical

Committees continued into 1989, and operational trials and final adjustments related to
implementation remained to be carried out in 1989-1990. In addition, much
administrative work on the standard had not even been initiated. The transfer of work on
GSM was therefore carefully staged in order to minimise disruption. Parallel structures --
(i.e., similar Technical Committees) -- were created in ETSI in order to ensure a smooth
transition. And, when the Groupe Spécial Mobile transferred from CEPT to ETSI in
1989, it preserved the same organisational structure and agenda for the GSM project.
Despite all of these measures, the transfer of GSM to ETSI translated into a considerable
delay in the introduction of GSM. ”The main effect was to make the GSM
recommendations subject to formal ETSI procedures, which required an extended period
during which specifications were subject to public inquiry (and possibly modification)
before they could be finally adopted” (Garrard 1998: 134).

ETSI not only took over the GSM project from CEPT but also assumed the mandate and
responsibility for developing further EU telecommunication standards (Drake 1994).
Henceforth, CEPT would limit its activities to matters of technical harmonisation. It is
not surprising that CEPT agreed to the transfer of its role in developing standards only
under ”intense pressure for changes in the procedures, focused through the European
Commission” (Garrard 1998: 134). It is also unsurprising that the foundation of ETSI
meant a break with former practices, and ushered in a new, more ‘open’ approach to
standards development, in which all relevant interests were included. ETSI and its
counterparts in the US (the T1 Committee, founded in 1984) and Japan (the
Telecommunications Technology Committee or TTC, founded in 1985) are ”regional in
scope and oriented towards the voluntary/consensus method of standards making”
(Hawkins 1995: 29).4 Because this approach ”calls for open and balanced participation in
standards committees”, ETSI, like its counterparts, has been ”virtually required to include
users from the start” (Hawkins 1995: 29).5

  Earlier in this text (in sub-sections 3.2.2 and 3.2.3.) the US and Japanese counterparts of ETSI were
identified as the TIA and the RCR, respectively. The references made here to T1 and RCR may therefore
be somewhat confusing. This problem arises due to the fact that a multi-purpose organisation such as ETSI
can have more than one ’functional equivalent’ in other settings. For example, as Hawkins et al. explain
elsewhere, TI – an abbreviation for ANSI T1 – is a committee of the American National Standards Institute
(ANSI). As such it responsible for the approval of ’voluntary’ standards developed by ANSI-accredited
organisations such as the TIA (Hawkins, Mansell, and Steinmueller 2000: 265). ETSI, however, combines
the functions of approval and development, which remain separate in the US standards development
  In this context, ‘users’ does not refer exclusively or even primarily to end-users. The author, whose work
deals with the development of technical standards in telecommunications, also excludes PTOs or PTTs
from the category of ‘users’. He considers that ”Most users can be accurately described as subscribers to a
level of service as defined by a monopoly public network operator (PTO)” (Hawkins 1995: 23). Of these,
he is mainly concerned with ‘intermediate- users’, rather than ‘end-users’. ‘Intermediate- users’ are
defined as follows:

              The ‘intermediate-user’ extends the functionality of the basic public network facilities. This
              may involve something as relatively simple as the provision of customised features by means
              of customer-owned equipment -- such as advanced private automatic branch exchanges and
              specialised data communication technologies. It could also involve something as complex as
              the interconnection of an entire private network to the public network. These service
              ‘enhancements’ may be confined to the user’s own internal requirements, but they may also

The ETSI membership consists not only of the large public network operators (PTOs and
PTTs) and established telecommunications equipment manufacturers, but also public
research organisations and new entrants into telecommunications service provision.
Moreover, the sheer number of members, now over 300, ”suggests that ETSI is reaching
well down to small to medium sized enterprises” (Temple 1992: 177). ETSI is also
‘open’ in procedural terms. The organisation’s emphasis on extensive public inquiry has
already been mentioned. In addition, ETSI has instituted a number of rules and
procedures aimed at the effective and reasonable resolution of disputes. These include:
precluding ‘national’ positions from arising within the Technical Committees; assuring
that the views of minority parties are heard and considered by the Assemblies; and
resolving disputes through an ‘indicative voting’ mechanism that requires members to be
present and enables decisions to be made on the basis of ‘best arguments’ rather than
entrenched positions (Temple 1992: 179). Rather than allowing individual members to
block decision-making processes by requiring unanimity, ETSI takes decisions on the
basis of “71 per cent weighted majority votes” (Bekkers and Liotard 1999: 113).6

To summarise, several innovations in procedure and organisational membership marked
the birth of ETSI. These, together with the regulatory powers that the European
Commission remained capable of exercising over ETSI decisions (see sub-section 4.2.1.),
ensured that standards developed within ETSI would serve the entire EU, as opposed to
any particular national industrial policy interests within it. These principles, together
with the EC’s project of liberalising telecommunications in the EU on the basis of
common standards that would consolidate and unify a ‘single market’ and strengthen the
EU’s competitive position in telecommunications, informed the further development of
GSM under the aegis of ETSI. Hence, for example, the MoU, under the direction of
ETSI, effectively placed manufacturers under essentially the same constraints as the
PTTs and PTOs: they would be required to serve the entire GSM community on equitable
terms and without discrimination (Bekkers, Duysters, and Verspagen 2000).

4.2.3. The MoU Group

 As explained at the outset of section 4.2 and discussed further in section 4.2.1., an
organisation responsible for administering the Memorandum of Understanding (MoU) on
GSM was first formed within CEPT in 1987, at the urging of the European Commission.
Later, and again at the suggestion of the Commission, the MoU Group relocated to
become part of ETSI, soon after its creation in 1987-88. The original signatories to the
Memorandum of Understanding (MoU) were the PTOs or PTTs of some 13 European
countries. By 1996, though, the list of signatories had grown to include 167 operators
from 103 countries and continued to expand rapidly (see section 1.3). By joining the

              be extended, in that they can be sold on to other parties. In practice, this can make a user
              resemble a supplier and vice versa. (Hawkins 1995: 24)
  The voting procedures are actually more complicated than this, since there are two voting structures with
different formulae for ’weighting’. Decisions on standards are made by national delegations, according to
the formula described above. Other decisions are based on weighted voting by all full members. A more
complete account of voting structures and weighting formulae is given in Bekkers and Liotard (1999: 113)

MoU Group, telecom operators accepted both a timetable and a set of regulatory
conventions concerning implementation of the GSM standard (see sub-section 4.2.1.).

The Memorandum of Understanding was based upon, and articulated, a “co-ordinated
and phased approach to procurement of network infrastructure” (Garrard 1998: 131). The
MoU also ensured that common policies would be formulated concerning a number of
issues. Many of these were commercial ‘service’ issues such as tariff principles,
procedures for billing, payment, accounting, and other problematic aspects of handling
‘roaming’ calls made in other countries, and so forth. In addition, the MoU was also
concerned with the further technical development of the GSM standard, and its
administration therefore sought to develop a consistent set of policies regarding the
Intellectual Property Rights (IPRs) associated with the standard. (Ibid.) In respect to the
last-mentioned activities, especially, the MoU Group demonstrated a great deal of
continuity with its origins in CEPT by proposing policies that clearly advantaged ‘users’
over ‘producers’. As one commentary remarks, “one must bear in mind here that the
CEPT members were mainly IPR users, not IPR creators” (Bekkers and Liotard 1999:

The commercial issues that the MoU Group had to contend with were settled largely by
resorting to the established practices and conventions that member organisations had used
for analogue networks. Beyond agreement that there should be basic registration fees, as
well as subscription and calling charges (billed to the caller), there was no attempt to
standardise tariffs among the various operators (Garrard 1998: 138). Roaming, however,
proved to be a more controversial matter. It was eventually decided to charge roaming
callers at local rates, and then transmit the call details and charges to their ‘home’
networks, which could then charge additional fees for ‘handling’ when presenting the
bill. Calls made to a mobile user who was ‘roaming’ internationally were even more
difficult, since they raised the issue of who should pay for the additional charges
associated with forwarding the call – caller or recipient?. It was finally decided that, in
spite of the ‘Calling Party Pays’ principle, callers should pay only the regular charge for a
call to a mobile on its home network and the recipients should pay the additional charges.
This was a discouraging decision for the mobile user, who could end up paying a great
deal for unwanted or unnecessary calls. Therefore, “it was essential to provide the option
for roaming users to limit use of their mobiles to outgoing calls while they were roaming”
(ibid.: 139).

In contrast to commercial or ‘service’ issues, issues related to infrastructure procurement
and the further technical development of the GSM standard turned out to be highly
problematic for the MoU group (Bekkers and Liotard 1999: 120 - 122). The
aforementioned attempt to develop a common policy on IPRs (Intellectual Property
Rights) encountered opposition from equipment manufacturers – especially, Motorola.
The MoU group, in its first draft procurement contract, proposed to give manufacturers
system licenses only if they would agree to secure the operators against any possible
patent infringements. A further condition was that manufacturers should also allow the
free use of any patents they held that were essential to the implementation of GSM
(Wilkinson 1991). All manufacturers accepted the first condition, though not without

complaint. However, while European manufacturers interested in the rapid market
development of GSM “grudgingly accepted” the second condition, the US-based
manufacturer, Motorola, bitterly opposed it (Garrard 1998: 140).

The reasons for Motorola’s opposition to the MoU’s policy on IPRs have been discussed
above, in sub-section 4.1.2. There, it was pointed out that Motorola’s main hope of
generating revenue from its heavy R&D investment in technologies incorporated in the
GSM standard resided in the ‘selective licensing’ of its patents, many of which were
crucial to GSM (Bekkers, Duysters, and Verspagen 2000). Firms based mainly within the
EU could, unlike Motorola, rely with greater confidence on other forms of participation
in the further development of GSM. But even among European manufacturers there was
“quiet outrage” regarding the IPR policy initially proposed for GSM under the auspices
of CEPT (Bekkers and Liotard 1999: 120). In effect, this policy meant that “IPR owners
could … be forced to give away the results of their research efforts without compensation
simply because their technology was selected for a CEPT standard” (ibid.). Motorola’s
determined resistance therefore sparked much a broader opposition among
manufacturers, initiating a conflict that took years and repeated attempts at reconciliation
to resolve (Cattaneo 1994: 64).

The conflict was eventually settled under the auspices of ETSI, which sought from the
outset to develop a ‘balanced’ approach to the relationship between standards and IPRs –
i.e., one that would not sacrifice public interests to property rights, or vice versa (Prins
and Seissel 1993). In the end, the European Commission’s position, that intellectual
property rights would have to be respected and compensated in order to sustain private
sector investments in research and development and to allow E.C. markets to benefit
from them, prevailed. The result was the approval and announcement, in 1993, of ETSI’s
new IPR Policy and General undertaking, which recognised the right of IPR holders to
claim revenues from their patents, but retained the principle of non-exclusive compulsory
licensing on ‘fair and reasonable’ to other ETSI members by means of a formula that has
been characterised as “licensing by default” (Bekkers and Liotard 1999: 121).

Thus, a compromise was finally reached whereby ETSI – and, by extension, the MoU
Group – retained effective control over the GSM standard, while manufacturers retained
limited claims to their IPRs. These claims were still severely limited, though, by the fact
that the original ETSI policy did not allow, for example, non-monetary considerations to
licensers – in particular, cross-licensing . However, this condition was soon annulled, in
response to a more extensive complaint filed with the EC Competition Authorities by the
US-based Computer and Business Equipment Manufacturers Association (CBEMA)
(Tuckett 1993). Subsequently, manufacturers were permitted to enter into bilateral cross-
licensing agreements with one another in relation to ETSI standards such as GSM
(Bekkers and Liotard 1999: 122 - 123). This was an arrangement which suited large firms
like Motorola well enough, although it raised steep entry barriers to firms that had no
valuable GSM-related (or other) patents to bargain with (Garrard 1998: 140). The
prospects for granting cheap licenses for the low-cost manufacture of GSM technology
outside of Europe were greatly reduced, if not eliminated altogether (ibid.).

4.3. Other Public-Sector Actors

While both the PTOs or PTTs and the SDOs played key roles in developing the GSM
standard, other public sector organisations were also important to this effort. Public
research organisations, such as universities and research institutes, in particular, played a
significant part in the development of second-generation of mobile telecommunications

The increased importance of public research organisations for the second generation of
mobile telephony stands in marked contrast to the case of first generation mobile
telephony. There -- for NMT, at least -- public research organisations did not make any
notable contributions.

4.3.1. Public Research Organisations

These points can be illustrated with reference to Ericsson’s growing involvement with
public research organisations in Sweden during the late 1980s and early 1990s.
According to McKelvey et al. (1998: section 9.2), the ‘Swedish competence build-up’
associated with the development of GSM depended at least partly on publicly funded
research involving collaboration between academic and industrial partners. Work on
digital standards in telecommunications formed part of the state-funded ‘IT4’ research
programme administered principally by NUTEK (formerly the National Board of
Technical Development). IT4 supported a number of specific projects that ran from 1987
to 1990, and ”Over time, universities were increasingly drawn into these projects”
(McKelvey, Texier, and Alm 1998: 50).

According to the same source, university programmes of teaching and research in
electronics and communications technologies in Sweden were also subject to growing
influence, not only from the Swedish PTO, Televerket, but also from Swedish private-
sector actors in GSM. These included both Ericsson and some smaller private-sector
actors involved in the manufacture of mobile telephone equipment in Sweden, such as
SRA (later renamed ERA, when it was acquired by Ericsson). Both Televerket and
Ericsson had participated in university faculty boards and financed specific professors.
For its part, ”SRA (later ERA) helped universities build test equipment” (McKelvey,
Texier, and Alm 1998: 50). Such co-operation continued into the 1990s and was
reinforced by further support from NUTEK for research programmes involving
university-industry collaboration, such as the 1993 - 1996 Research Programme on
Telecommunication and Service (NUTEK 1997).

Televerket, Ericsson and other actors benefited from the development of competencies in
public research organisations in at least two ways. On one hand, an expansion of
research capabilities beyond existing ‘in house’ resources had been necessary for them to
master the technological discontinuity posed by the transition from analogue to digital
mobile telephony. On the other hand, building up the competencies of the universities, in
particular, ”seems to [have been] particularly important for the recruitment of engineers,

which has been a critical need for Ericsson in the 1990s at their fast pace of expansion”
(McKelvey, Texier, and Alm 1998: 51).

A similar picture can be drawn for the EU. There had, of course, been no concerted effort
at the EU level to involve public research organisations in the development of first-
generation mobile telephony. With the second generation, however, came the first efforts
at bringing about closer university-industry collaboration in this area. Most of these
initiatives appear to have been taken after the publication in 1987 of the EC Green Paper
on Telecommunications (Commission of the European Communities 1987). Given that
the GSM standard had already been successfully defined for the most part by 1987, the
majority of the EU-funded research programmes that were initiated at that time were
oriented more towards the eventual development of third-generation technologies
(DaSilva and Fernandes 1995). However, certain collaborative research intiatives
supported by the EU dealt with problems whose resolution could contribute to the
‘evolution’ of GSM in its second phase of development.

One such initiative has been COST (Co-Opération européenne dans le domaine de la
recherche Scientifique et Technique), which was closely associated with mobile-related
projects during the 1980s. COST, a co-operative arrangement founded in 1971 and
supported and hosted by the EU, is not an EU-controlled organisation or research
programme. Rather, it has an independent basis as a European intergovernmental
agreement for promoting research collaboration, providing a multilateral framework for
co-operation aimed at improving the quality of European research in a variety of fields
(Nissinen and Niskanen 1999: 15). COST’s various programmes or ‘actions’ consist of
basic or pre-competitive research on problems of Europe-wide interest and are directed
primarily towards public research organisations. However, these ‘actions’ often include
business enterprises and other private sector organisations as participants, and one of the
general objectives of COST has been to forge collaborative links between industry and
academia (European Commission 1996: 1).

Telecommunications has been one of the most heavily subscribed domains of COST
collaboration (European Commission 1996: 4). One of the most successful COST
‘actions’, moreover, was directly linked with GSM: ”COST 207 made a major
contribution to the development of the harmonised Pan-European GSM Public Digital
Mobile Radio System, which has spread not just across Europe but also beyond Europe”
(Nissinen and Niskanen 1999: 28). In addition, other COST ‘actions’ also supported the
further evolution of GSM and mobile telephony. COST 227 sought ”to define systems
for mobile communications based on the integration between satellite and terrestrial
networks”, and COST 231 studied ”digital transmission techniques and propagation
aspects” (DaSilva and Fernandes 1995: 16).

The results of these projects were, in due course made available to ETSI, which
incorporated them into the standard development process. As a consequence of ETSI’s
strong reliance on co-ordinated public-sector research initiatives such as the COST
actions mentioned above, its ongoing standardisation projects – with GSM and DECT

referred to in this context as prominent examples – have been characterised as “RTD&D
projects in all but name” (Hawkins, Mansell, and Steinmueller 2000: 255).

5. Private-Sector Actors

As already indicated, in section 4 (see, for example, sub-sections 4.1.1 and 4.1.2), some
of the most important private-sector actors in the development of GSM were the
‘national’ telecommunications equipment manufacturers that co-operated with the
European PTOs or PTTs submitting prototype technologies for the GSM standard. Most
of these firms -- and especially the Nordic firms, Ericsson and Nokia -- subsequently
became major producers of equipment for the GSM standard. As participants in the
phenomenal growth of GSM, these firms gained a significantly increased share of the
world market for global telecommunications equipment. However, other private sector
actors also played an important part in the development of GSM. In particular, new
service providers emerged in national and international markets, posing a competitive
challenge to the PTOs/PTTS and presenting opportunities for further innovation within
the GSM trajectory. Moreover, the PTOs and PTTs themselves also underwent
privatisation in some cases. Thus, some of them eventually participated in the further
development of GSM as ‘private sector actors’, even though they began their
involvement as public sector organisations.

5.1. Equipment Suppliers

In section 4.1 it was noted that the development of GSM technology involved extensive
collaboration between ‘national’ European telecommunications equipment manufacturers
and national PTTs/PTOs (see sub-section 4.1.1). This was especially the case in
connection with the submission of prototype technologies for the GSM standard (see sub-
section 4.1.2.). However, subsequent interaction was also important, especially for major
Nordic equipment producers such as Ericsson and Nokia. Their collaborative activities
took place within the framework of a loosely-knit ‘alliance’ among Nordic firms -- Elab,
Ericsson and Nokia -- and the PTTs/PTOs of Finland and Sweden. Ties between firms –
not all of them Nordic – were also important in this context. The Nordic alliance, which
prevailed in the selection of the prototype technology for GSM, did so under the effective
leadership of the Swedish PTO, Televerket, later privatised to become Telia.7
Televerket’s TDMA system was chosen by CEPT to become the basis for GSM, creating
an initial advantage for Nordic producers. Subsequently, Ericsson, together with
Televerket, developed and tested the first prototype of a full GSM system, thus
consolidating their early technological ‘lead’. Nokia also benefited from the GSM
decision., as did other firms in the alliance.

 Telia has not actually been privatised -- at least not fully. Rather, what has transpired could be described
as a first step towards privatisation. Telia remains under state ownership, but has been transformed from a
public enterprise (Affärsverk) to a joint-stock, limited liability company (Aktiebolag, abbreviated AB).
(Karlsson 1998: Chap. 3).

5.1.1. Ericsson

During the late 1980s, under the auspices of the Swedish ‘IT4’ research programme
mentioned in section 5.3.2, Ericsson co-operated with Televerket in the construction of a
GSM system prototype (McKelvey, Texier, and Alm 1998: section 9.2). This
collaboration, conducted mainly at Ericsson’s R&D facilities, proceeded from initial tests
of the NMT 900 system, through construction of a trial GSM system, to development of
key components for GSM. Ericsson was able to draw upon related academic research, as
well as the resources of Televerket. At the completion of the project, Ericsson and
Televerket could demonstrate the capacity to simulate and test an entire GSM system,
leading to the result that their system proposals became accepted as the basis of GSM in
Europe (Lilliesköld 1989: 35). This accomplishment, however, was not simply achieved
through Ericsson’s collaboration with Televerket. It also depended strongly on co-
operation and the strategic exchange of technology between Ericsson and other firms.

As discussed previously (in sub-section 4.1.2), Televerket and Ericsson could not claim
exclusive authorship of the prototype technology that was eventually selected as the
basis for the GSM standard. Although the MoU Group responsible for the development
of GSM, first within CEPT and later within ETSI, sought to appropriate the essential
patents associated with the GSM standard, this move was resisted by the manufacturing
firms, led by Motorola. This dispute was settled with the outcome that these firms
remained able to cross-license their patents to others. (See sub-section 4.2.3.) Subsequent
publication of information concerning the essential GSM patents by ETSI has shown that
of 132 essential patents in all, the largest share (50%) was claimed by Motorola. The
second largest share (16%) was claimed by AT&T, and that Bull and Phillips both
claimed the third largest shares (8% each). (Bekkers and Liotard 1999: 123) Since none
of the firms named here are based in the Nordic countries, this record indicates that at
least 82% of the essential patents for the GSM standard were of non-Nordic origin.
Moreover, it implies that major Nordic equipment producers such as Ericsson had to
engage in extensive cross-licensing in order to acquire a full system capacity with
respect to GSM.

During the early 1990s, Motorola, the holder of fully half of the essential patents for the
GSM standard, entered into cross-licensing agreements with four other firms. In addition
to Ericsson and Nokia, Motorola also concluded such agreements with Siemens and
Alcatel (Funkschau 1993). Interestingly, none of these firms claimed any large share of
the essential patents for GSM. This observation strengthens the argument that Motorola
“imposed a market structure by conditioning exclusive cross-licensing agreements with a
number of other parties on the market” (Bekkers, Duysters, and Verspagen 2000). By this
means, Motorola gained significant revenue from the growth of a market for GSM
infrastructure and terminal equipment from which it might otherwise have been excluded.
The European-based firms with which Motorola concluded such agreements were also
beneficiaries of this arrangement. As revealed in Table 5.1, below, all of these firms
quickly became major suppliers for the GSM standard and, as a group, dominated this

Table 5.1: Estimated Suppliers’ Market Share of the 33 Largest GSM Networks in
Europe, December 1996, plus World-Wide Market Share of GSM Terminals during

Supplier       Score           Market         Score Base      Market         Market
               Switching       Share          Stations        Share Base     Share
                               Switching                      Stations       Mobile
Ericsson       10,297          48%            7,978           37%            25%
Siemens         4,426          21%             325             2%             9%
Nokia           3,086          14%            4,617           22%            24%
Alcatel         2,228          10%            2,084           10%             6%

Source: Table 3, Bekkers and Liotard (1999)

Table 5.1 demonstrates that even though GSM was conceived as being, in principle, an
‘open’ system, at least among all telecom operators and suppliers that were signatories to
the MoU, the licensing of essential patents conferred an important competitive advantage
on certain suppliers, including Ericsson. The strategic value of cross-licensing was
clearly apparent at a very early point to Ericsson, as well as to the other firms that
participated in these agreements. Ericsson, in particular, appears to have used access to
its powerful AXE switching technology as a basis for negotiating such agreements in the
case of acquiring key GSM patents (Bekkers, Duysters, and Verspagen 2000; Glimstedt
2000: 8).

The AXE switch, which had already provided Ericsson with entry into the North
American market for ‘first generation’ mobile telephony, had already proven to be a vital
infrastructural complement to the systems being developed by US-based producers such
as Motorola (Meurling and Jeans 1994: 77),as well as leading to Ericsson’s development
as an important supplier of mobile telecommunications infrastructure in the US
(Meurling and Jeans 1995: 179). Moreover, further adaptation of the AXE switch
required not only the transfer of technology from Ericsson but also considerable R&D
investment. These observations make understandable Ericsson’s initial reluctance to
agree to inter-operability of switches and radio-subsystems as an essential design feature
of GSM – a decision originally made by the Groupe Spécial Mobile in order to ensure
the interchangeability of switches produced by different suppliers and, hence, competitive
sources of supply for switches. (See the discussion in sub-section 2.1.3.)

Having used its AXE technology as a basis for entry into the first generation of moblile
telecommunications, via the development of NMT, Ericsson later used its commanding
position in switching for the acquisition of other technologies that would give the firm an
advantageous position in relation to GSM in the second generation of mobile

telecommunications. However, Ericsson did not concentrate exclusively on GSM. Rather,
the firm’s strategy was to pursue simultaneously all of the three major international
standards for the second generation (McKelvey, Texier, and Alm 1998: section 9.3). This
meant further and even more extensive reliance on collaborative strategies such as cross
licensing. At the same time, Ericsson sought to build up a specialised but comprehensive
‘systems’ capability in mobile telecommunications through various means. Such devices
included, for example, acquiring the few competing domestic firms in mobile
telecommunications, to exploit more fully its former competitors’ complementary
strengths in infrastructural components such as radio base stations (ibid.: section 9.1).

During the 1990s, Ericsson responded to the rapid growth of GSM and mobile
telecommunications in general with a series of broad measures intended to broaden its
competency base. To accomplish this, Ericsson pursued a three-point strategy: 1) to
increase greatly its R&D expenditures; 2) to keep only ‘core’ activities within the firm
(i.e., ‘in-house’); and 3) to enter into collaborative relations (such as strategic alliances
and joint ventures) both with other equipment suppliers and with network operators and
service providers in strategic national markets such as those of France and Germany.
(McKelvey, Texier, and Alm 1998: section 9.3)

Ericsson’s dramatically increased expenditures on R&D during this period have been
well documented. According to one recent account (Granstrand, Patel, and Pavitt 1997),
Ericsson’s R&D spending approached 20% of sales in the 1990s. Even earlier, during the
1980s, the total number of engineers employed by the firm had risen by 82% (ibid.).
However, it is less clear that Ericsson’s R&D expenditures reflected an increased
concentration on a restricted set of ‘core competencies’. Rather, there was an increased
diversity of competencies in engineering:

            The traditional core competence in electrical engineering increased by only
            32%, while mechanical engineering grew by 265%, physics by 124%, and
            chemistry by 44%. Additional engineering categories were added (e.g.,
            computer science), and no broad category of engineering competence was
            scrapped. (Granstrand, Patel, and Pavitt 1997)

Ericsson’s collaborative efforts, in particular, yielded commercial success while at the
same time demonstrating the superiority of Ericsson’s switches and radio base stations to
those of alternative suppliers. For example, in Germany, one of the most important initial
markets for GSM, the newly licensed operator/service provider, Mannesman Mobilfunk,
relied heavily on Ericsson during the early 1990s to construct ”by far the largest GSM
network that anyone had built” (Garrard 1998: 258). Mannesman ordered its initial
infrastructure form Ericsson, which ”rapidly became recognised, along with Nokia, as the
surest way to achieve trouble-free roll-out” (ibid.). In the second phase of its roll-out,
Mannesman replaced less satisfactory switches and radio base stations from other
suppliers ”with their Ericsson equivalents” (ibid.)

Ericsson’s international collaborations also enabled the firm to gain access to new
markets and to build market share in areas of strategic importance. The latter function of

collaboration, especially, was important to Ericsson complementing its strategy of
focusing more closely on ‘core’ business activities such as mobile telephony.

Historically, Ericsson had a well established record of relying upon foreign subsidiaries
and joint ventures as a means of establishing a presence in distinct geographical markets
(Fridlund 1998). However, many of the firm’s overseas holdings took the form of
‘branch factories’ manufacturing an increasingly wide range of telecommunications
equipment (Wells and Cooke 1991: 96 - 97). In the late 1980s, though, the increasing
importance of software costs as a proportion of total system costs (then approaching
75%), had begun to undermine the profitability of existing branch manufacturing

The restructuring initiated in response to these conditions involved the shedding of recent
acquisitions in cables and computers, in order to build up core competencies in
telecommunications (Dixon 1988). At the same time, Ericsson introduced a new,
‘modular’ approach to system design, in which market adaptations of ‘core systems’ were
made locally (Meurling and Jeans 1995). In this context, Ericsson’s branch factories
were either closed down or remade into regional ‘competence centres’.

Ericsson had continued to use joint ventures as a means of establishing itself in key
markets that were otherwise difficult to enter -- as, for example in its partnership with the
UK firm, Thorn-EMI in the equity joint venture Thorn-Ericsson, which Ericsson took
over in 1988 (Dodsworth 1988). But with increasing liberalisation of formerly protected
national markets, joint ventures and strategic alliances became more important as a
means of gaining market share in specific product markets. In mobile telephony, Ericsson
had used alliances with Orbitel in the UK and Matra in France to build up a 40% share of
the world market in radio base stations. In the late 80s, though, the firm still remained
weak in the market in the market for terminal equipment, with only 14% of the world
market and 3% of the US market (Wells and Cooke 1991: 99). A strategic alliance was
therefore formed with US-based General Electric in order to increase sales and capture
economies of scale in terminal equipment, such as hand-sets (Taylor and Dixon 1989).

Ericsson, together with Nokia, eventually enjoyed considerable success with the sales of
terminal equipment. ”By the end of 1995, just three manufacturers -- Motorola, Nokia,
and Ericsson -- still shared about 75% of the GSM terminal market, with only a couple of
percentage points between them” (Garrard 1998: 140). For Ericsson, this success
reflected a build-up of ‘core competence’, through R&D and other means, but it also
rested on collaborative foundations.

Concentration of the supply of GSM hand-sets was due, in part to the negotiation among
manufacturers of bilateral agreements for the cross-licensing of patents. (See the
extensive discussion above). As discussed previously (in sub-section 4.2.3.), these
agreements were made in response to the procurement policy of the MoU group, which,
in addition to imposing stiff barriers to market entry, sought an extensive relaxation of
normal patent rights from manufacturers as a condition for the award of systems
contracts. These demands were only grudgingly accepted by European manufacturers,

and steadfastly resisted by US-based Motorola.8 Eventually, cross-licensing agreements
between manufacturers became accepted as a compromise solution.9 But this solution
favoured concentration and oligopoly: ”This arrangement was satisfactory for the larger
players, which had their own IPR to use in negotiations, but less acceptable for the
smaller companies, which were less likely to have such an advantage” (Garrard 1998:

The business strategy pursued by Ericsson was successful, not only in commercial terms,
but also in terms of redefining and broadening the firm’s competency-base. Ericsson has
since been cited as a leading example of increasing ‘technological diversity’ in large
corporations manufacturing complex and technically sophisticated products (Granstrand,
Patel, and Pavitt 1997). In the 1990s, Ericsson clearly made the development of superior
expertise in mobile telecommunications a centrally important focus of activity. However,
this had meant widening the firm’s competencies, not narrowing them.

This tendency was well-defined, even before the 1990s. During the post-war period,
Ericsson had progressed from an initially narrow focus on telecommunications equipment
and related technologies (electro-mechanical switching and cable transmission) to master
computerisation and digital switching technologies, later combining them with radio
transmission technologies in order to develop the NMT system (Fridlund 1998). In the
early 1990s, this tendency of competence-enhancement through technological diversity
was intensified with the development of GSM. The firm’s leading products became more
technologically complex and its competence base accordingly came to include a greater
number of technological competencies (Granstrand, Patel, and Pavitt 1997).

The increased diversity of Ericsson’s expanding competency base, as well as the firm’s
growing tendency to acquire new technologies from external sources and the expanding
scale of its R&D efforts is shown in Table 5.2. This table compares Ericsson’s
competency-base when GSM was first implemented with its earlier states corresponding
to the development of the NMT 450 and NMT 900 systems. Competency development
related to two successive new cable transmission systems (coaxial and fibre-optic cable)
is also described.

One significant outcome of Ericsson’s narrowed business focus and broadened
competency-base was that the firm developed what could be termed ‘system competency’
in mobile telecommunications -- i.e., the capacity to design, build, market and support,
entire systems for the operation of mobile telecommunications networks. Originally,
Ericsson was only a supplier of certain components for PTOs or PTTs that had effectively
acted as system integrators (and architects) in their capacity of ‘network operators’.
However, through its involvement in the first two generations of Nordic mobile
  In this context, Motorola, despite being based in the US, was considered by the MoU group to be a
corporate citizen of the European Union. This recognition of Motorola as an ”honorary European” was ”by
virtue of its extensive R&D activities in several European countries” (Garrard 1998: 140).
  This compromise was not reached easily, nor did it establish a precedent by which future conflicts
between standards and intellectual property rights (IPRs) in telecommunications could be minimised. For a
detailed account of the issues and how they were resolved in the case of GSM, see Bekkers and Liotard

telephony, Ericsson had acquired the capability to develop comprehensive solutions
addressing all three of the main technology areas in mobile telecommunications -- i.e.,
radio base stations, switches, and terminal equipment.

Table 5.2: Technology Accumulation in Ericsson’s Product Generations

Product     No. of     No. of     Total      No. of     R&D       % of       Main      No. of
and         Old        New        Number     Obso-      Costs     Techno-    Techno-   Patent
Product     Tech-      Tech-      of Tech-   lete       (base =   logies     logical   Classes
Genera-     nologies   nologies   nologies   Tech-      100)      External   Fields    (e)
tion        (a)        (b)                   nologies             -ly Ac-    (d)
                                             (c)                  quired

1. NMT -
450      n.a.          n.a.       5          n.a.       100       12         E         17

2. NMT -
900      5             5          10         0          200       28         EPM       25

3. GSM      9          5          14         1          500       29         EPMC      29


1. Coaxi-
al        n.a.         n.a.       5          n.a.       100       30         EPM       14

2. Optic-
al          4          6          10         1          500       47         EPMC      17

n.a. = not applicable.
(a) No. of technologies from the previous generation.
(b) No. of new technologies, compared to previous generation
(c) No. of technologies obsolete from previous generation.
(d) ‘Main’ = >15% of total engineering stock. Categories are: E = electrical; P = physics;
M = mechanical; C = computers.
(e) Number of international patents (IPC) at four-digit level.
Source: Derived from Granstrand et al. (1992).

At least one recent study of Ericsson’s development into one of the major producers of
mobile telecommunications equipment has concluded that the firm’s competitive
advantage has increasingly depended on the acquisition and effective utilisation of this
kind of ‘system competence’. ”Thus, it has been important that Ericsson have

competencies in all three major technology areas but also in order to integrate a system
and identify opportunities it has been important that competencies have been spread
within the company” (McKelvey, Texier, and Alm 1998: section 9.4).

5.2. Service Providers & Network Operators

New network operators and service providers were also important private-sector actors
involved in the development and diffusion of GSM. In contrast to the innovative roles
played by the PTOs or PTTs and the major equipment producers, the main contribution
of these new entrants was not to introduce ‘radical’ technological innovation but rather to
create new markets and expand existing ones. However, as indicated by the discussion of
PCNs in sub-section 3.1.2, this activity also involved ‘incremental’ technological
innovation to some extent.

5.2.1. A New Policy Regime in the European Union

As noted in section 4 (see, especially, sub-section 4.4.1.), GSM was introduced more or
less simultaneously with the advent of a new telecommunications policy regime in the
European Union. The overarching goal of EU policy, as indicated in the European
Commission’s 1987 Green Paper, was to deregulate the telecommunications sector in
order to create a competitive common market in telecommunications equipment and
services (Commission of the European Communities 1987).

CEPT’s development of GSM coincided with the announcement of this policy.
Moreover, as noted earlier (in sub-section 4.2.1), it was regarded by the European
Commission as a welcome departure from the non-competitive and non-innovative status
quo in European telecommunications. Subsequently, GSM was effectively ‘taken over’
by the EC and moved to ETSI, which steered its further development in the directions
indicated by EU policy. (See sub-section 4.2.2.) In this context, GSM became ”the
spearhead for European policy to liberalise telecommunications” (Garrard 1998: 247).

Although mobile telephony was deliberately not addressed in the EC’s 1987 Green
Paper, it later became the subject of a 1994 Green Paper on a Common Approach to the
Field of Mobile and Personal Telecommunications in the European Union (Commission
of the European Communities 1994). This document contained five main proposals that
subsequently engendered much controversy and were greatly revised. Their common
purpose was to complete the liberalisation of European telecommunications markets and,
thereby, to “remove the barriers to further development” of mobile telecommunications
(ibid.: 37). The specific proposals read in part as follows:

        1) abolishing remaining exclusive and special rights in the sector, subject
           where required to the establishment of appropriate licensing conditions;

            2) removal of all restrictions on the provision of mobile services both for
               independent Service Providers and direct service provision by mobile
               network operators;

            3) full freedom for mobile network operators to operate and develop their
               networks for the purpose of the activities provided for in their license or

            4) unrestricted combined offering of services via the fixed and mobile
               networks, within the overall time schedule set by Council Resolution
               93/C213/01 of 22 July 1993 for the full liberalisation of public services via
               the fixed network;

            5) facilitating pan-European operation and service provision.

                (Commission of the European Communities 1994: 37 - 38)10

One proposal (number 2) concerning the use of separate service providers (along with
direct service provision) by network operators was later dropped. Another (number 3),
concerning the right to self-provision by operators of fixed links for network connection
was considerably weakened, and the debate on other proposals continued without
reaching consensus. Eventually, the European Commission took unilateral action and in
January, 1996, issued a directive based on its two most crucial proposals. The directive
required that: 1) mobile services must be competitive, with multiple GSM licenses in
each member state, and 2) mobile operators should be able to construct their own
infrastructures for transmission. (Garrard 1998: 243 244) Subsequently, EC
implementation of this directive facilitated the gradual opening of national markets to
new network operators and service providers.

GSM, as developed by CEPT, was originally the product of the ‘old regime’ of protected
national telecommunications markets and public sector monopolies exercised by the
PTOs or PTTs. But the new standard in mobile telephony also became closely identified
with the EC’s drive to open the telecommunications sector to competitive market forces
during the 1990s, once GSM had ‘migrated’ to ETSI.

Arguably, both the ‘old regime’ and the liberalisation policies that followed it contributed
to the market success of GSM -- in different, though complimentary ways. CEPT, as a
monopolistic organisation comprised of monopolists, was able to achieve agreement on a
common technical standard for digital mobile telephony in Europe, and this
‘harmonisation’ made possible the consolidation of a huge initial market for GSM.
However, the entry of new network operators and service providers into the
     Emphasis in the original.

telecommunications sector under the banner of ‘liberalisation’ was what made possible
the wide diffusion of GSM within this large potential market.

For these reasons, the EU policies contributing to the success of GSM have been
discussed in terms of a ‘parallel path’ that maintained an almost ideal -- and, perhaps,
non-replicable -- balance between co-ordination and competition. ”There is a high level
of interaction -- and interdependence -- between these two approaches; harmonisation can
be considered a necessary condition to achieve true liberalisation and the breakdown of
(technical) trade barriers” (Bekkers and Liotard 1999: 111).

5.2.2. ... and in Sweden

Sweden, during the first period of GSM’s diffusion on the market, was not a member of
the Eropean Union and was therefore not directly subject to early EU policies regarding
the liberalisation of telecommunications. However, by the time Sweden finally joined the
EU, in 1995, it had already implemented many of these policies. As observed by the
OECD, Sweden became one of the leading European countries in the liberalisation of
fixed telecommunications networks that occurred during the 1990s. Together with a few
other European countries, such as Finland, Sweden ”went beyond the initial competition
policies pioneered in the 1980s” by the U.S., New Zealand and Japan (OECD 1999a:
11). There was a similarly rapid liberalisation consistent with EU policy directions in
Swedish mobile telecommunications.

The early liberalisation of both the fixed and mobile telecommunications networks in
Sweden was due to several related reasons. First of all, the Swedish telecommunications
sector had remained relatively un-regulated. The Swedish PTO, Televerket, had never
held a legal monopoly over telecommunications -- even though it had established a de
facto monopoly during the postwar period. Televerket had also exercised certain
regulatory powers over, for example, terminal equipment for the fixed network and the
allocation of radio frequencies for mobile network operators. (Karlsson 1998: Chap 5,
304) Nonetheless, there had never been legal barriers to establishing, e.g., new mobile
telecommunications networks in Sweden.

Second, the Swedish government had begun to discuss liberalisation at a very early point,
and had taken some preliminary steps in this direction during the 1980s. Throughout this
period, the discussion of liberalisation in Swedish telecommunications referred to
positive examples provided by the US and the UK. Towards the end of the decade it was
also increasingly conducted with a view to ‘harmonisation’ with the policies of the
European Union, such as those proposed in the 1987 Green Paper (Andersen and
Eliassen 1993; Petersson and Söderlind 1992: 63 ff, 214f). A government bill to limit the
monopoly powers of Televerket, and to ensure fair competition in the non-monopoly
sector of fixed voice telecommunications was introduced in 1980 (Karlsson 1998: 171).
During the 1980s, new competitors to Televerket’s mobile services -- Comvik, especially
-- made mobile telecommunications an ‘exception’ within the monopoly sector and
continued to contest the regulatory decisions of Televerket regarding frequency

allocations and other aspects of mobile telecommunications infrastructure. As a result,
open competition in mobile telecommunication infrastructure was not only well
established in practice by 1988 but also sanctioned by government policy that was well
on its way to becoming law (ibid.: Chap. 5, 284-285).

Third, the policies initiated at the very beginning of the 1990s included the separation of
regulatory decision-making from the regular business operations of Televerket. The
establishment in 1990 of a National Telecommunications Council (Statens Telenämnd, or
STN) to oversee this structural separation was soon followed, in 1992, by the
establishment of the National Telecommunications Agency (Telestyrelsen).
Telestyrelsen’s inauguration not only spelled an end to ”the increasingly problematic
double role of Televerket in the area of frequency management” but also marked the
creation of an independent telecommunications regulator capable of ensuring fair
competition in the non-monopoly telecommunications sectors, which now included the
mobile sector (ibid.: Chap. 5, 285 - 288).

Fourth, Swedish legislation enacted EC policies regarding telecommunications, even
before Sweden joined the EU. Sweden’s new telecommunications law, completed in
1993, not only consolidated the legal basis for the structural separation of regulatory
activities from the business operations of the Swedish PTO, Televerket. It also
introduced a number of other measures aimed at strengthening competition in
telecommunications. These measures ensured, among other things, implementation of
the EC policy directive on Open Network Provision (ONP). Based on the 1987 Green
Paper, ONP aimed at creating an open, competitive, and non-restrictive Europe-wide
market for telecommunications network operators (Braugenhardt and Kaaman 1991).
Sweden’s new telecommunications law was finalised in 1993. It was ”formulated to
comply with the European Community legislation” and it was also accompanied by a
modified Radio Transmission Law, under which frequency management and the
licensing and regulation of mobile networks passed from Televerket to Telestyrelsen
(Karlsson 1998: Chap. 5. 296, 298 - 301). .

For these several reasons, Sweden’s mobile telecommunications market was completely
open to competition from new entrants by the early 1990s. Shortly after the GSM system
was introduced, the oldest Swedish mobile telephone operator, Televerket’s ‘Radio’
department, was re-organised as a competitive business into the specialised subsidiary,
Telia Mobitel (Holst 1994: 189). It also met some competition from two new operators.
The first one was Comviq GSM AB, which entered the market in September, 1992
(Lernevall and Åkesson 1997: 571). At the same time Europolitan (owned by NordicTel
AB) also started to offer a GSM service.

5.2.3. Telia Mobitel

As observed previously (see,eg., sub-section 4.1.2.), the Swedish PTO, Televerket, had
clearly dominated the Swedish market for mobile telecommunications prior to the 1990s.
One reason for this was that Televerket had been able to position itself as the only
operator of NMT networks in Sweden. The only competitor, Comvik, had introduced a

different type of system, using network infrastructure imported from the USA and a
standard incompatible with NMT. In addition, Televerket, in its regulatory capacity, had
greatly restricted the radio frequency allocations available to Comvik, rejecting many of
its applications for additional frequencies.11 (Karlsson 1998: 242 - 244). For these
reasons, Comvik had never been a direct competitor to Televerket in analogue mobile
telecommunications and its share of the market remained low -- at less than 3% in 1990
(Lernevall and Åkesson, 1997: 565, 575). The Comvik network reached its capacity limit
of about 18,000 subscribers in 1990. At the same time, Televerket’s two NMT systems
had a combined number of 500,000 subscribers. (Karlsson 1998: 250) Telia Mobitel
could expect continued growth in demand for both analogue and digital mobile

Even before the introduction of GSM by Televerket/Telia Mobitel in 1992, Sweden’s
mobile telephone market had the highest penetration rate in Europe -- about 7%. Market
growth continued steadily, due to a combination of high quality service provision and low
tariffs. Sweden’s market was very dynamic, compared to the UK market, which some
observers thought to be saturated at a penetration rate of just over 2% in 1990. The
density of users in Sweden was three times that of the UK, and one of the main reasons
was that prices had been kept low, due to a ‘public service’ orientation on the part of
Televerket and other Nordic PTOs or PTTs. Sweden’s fixed subscription rates were
much lower than the UK’s, and charges were about half what they were in the UK.
(Garrard 1998: 188) For these reasons, competition in the market created by the
introduction of GSM under a ‘liberalised’ regulatory regime followed a somewhat
different pattern in Sweden than it did in the UK.

Due to the exceptionally high level of market penetration in Sweden, there was no similar
concern to that in the UK with developing a ‘mass market’ for GSM, based on a high-
capacity ‘consumer’ variant of the basic standard and technology for digital mobile
telephony. (See the discussion of PCN and DCS1800 in sub-section 3.1.2.) Conseqently,
the DCS1800 standard was not introduced at a particularly early point in the diffusion of
GSM in Sweden, and it did not ‘lead’ development of the GSM market in the same way
that PCN did in the UK. GSM was initially marketed to business users by Televerket
Radio/Telia Mobitel, which continued at first to rely primarily on the NMT 900 standard
to serve the consumer market. In 1994, Telia Mobitel’s exclusive NMT900 network still
had more subscribers than its new GSM network, though the latter was quickly catching
up, having experienced rapid expansion since opening in 1992 (Mölleryd 1997: 45). At
that time, the NMT900 network remained capable of considerable expansion, though
Telia Mobitel planned an eventual transfer of NMT900 frequencies to GSM (Garrard
1998: 265 - 266).

Competition in the ‘open’ GSM market from the new entrants, Comviq and Europolitan,
eventually forced Telia Mobitel to take a more profound interest in marketing GSM to
consumers. By 1995 most of Telia’s growth in mobile telephony was based on GSM,
and the proportion of non-business subscribers had steadily increased. Telia Mobitel,

 In one instance that later became notorious, Comvik’s application for frequencies in the 900 Mhz band
was refused because Televerket planned to launch its NMT 900 system in this band.

responding to the initiatives of its competitors, intensified marketing and commision sales
strategies, and also introduced several different kinds of subscriptions, which ”produced a
division of one private and one business market” (Mölleryd 1997: 45).

Due to the rapid growth of the GSM market in Sweden, which was accelerated by such
measures, Telia also applied, along with other Swedish GSM operators, for a license to
establish and operate a high-capacity DCS 1800 network. These licenses were granted in
1996 (Holst 1997: 110). However, the licenses granted to the new mobile operators were
‘dual band’ licenses that permitted them to open DCS1800 networks only after their
existing GSM networks had reached full capacity. In contrast, Telia Mobitel and a new
‘fixed network’ operator, Tele 8, were granted permission for an immediate start-up of
DCS1800 networks. (Garrard 1998: 268) In Sweden, therefore, it was primarily the
incumbent operator, Telia Mobitel, that pioneered the mass-market variant of GSM,
rather than new entrants as had been the case in the UK. (For the UK case, see sub-
section 3.1.2.)

At the same time that it began operating a DCS1800 network, Telia Mobitel also sought
to expand its share of both ‘business’ and ‘private’ markets by offering DECT (discussed
in sub-section 3.1.1.) as an added service to mobile users, enabling them to make cheaper
calls at fixed line prices from either their homes or offices. The technological innovation
that made this competitive strategy feasible was the dual-band hand-set, which was
developed by Ericsson to accommodate both the 900 MHz and MHz frequencies, but
could also be used to log onto cordless DECT terminals. (Holst 1997: 108) Relying on
both improved marketing and technical innovation, Telia Mobitel was able to improve its
share of the GSM market at the expense of its competitors during the mid-1990s. By
1997 it had re-emerged as the dominant Swedish GSM operator, controlling 54% of this
market (Holst 1999: 114).12

5.2.4. Comviq

Telia Mobitel’s main competitor, the ‘new entrant’ Comviq, was initially more
aggressive about marketing GSM to consumers. Although it was affiliated with the
newly established ‘fixed network’ operator, Tele 2, Comviq (introduced as Comvik in
sub-section 5.2.1), as explained above (in sub-section 5.2.2.), had been kept in a
disadvantageous position in, and had only obtained a minimal share of, the market for
analogue mobile telephony in Sweden. This was partly due to Comvik’s initial
technological choice to utilise a standard other than NMT 450 for its first network. It was
also partly due to restrictions imposed on further network development by the regulatory
decisions that had been made by Televerket as an allocator of radio frequencies.

Comviq’s response to having been effectively marginalised in the analogue market was to
decide, during the late 1980s, that GSM should be the main focus of its competitive
strategy during the 1990s (Karlsson 1998: 250). Comviq first applied to Televerket for a
reservation of frequencies in order to establish a GSM system in 1987. The application

  The measure of of market share here is based on the revenue of operators, rather than the number of

was eventually successful, thanks to government intervention. After obtaining a GSM
license by signing the MoU in 1990, Comviq purchased equipment from US and German
suppliers (Motorola, Digital Microwave, and Siemens) to establish its GSM network
(ibid.: 273 - 279).

Comviq’s explicit aim was to capture about 50% of the emerging GSM market. It
proposed to do so by means of a sales-oriented strategy emphasising low-cost mobile
telephony. After its first year of operation, Comviq turned increasingly to the consumer
market, and in doing so employed many of the marketing strategies that had been used to
‘grow’ the mass market for PCN in the UK. (See sub-section 3.1.2). Along with its
competitors, Telia Mobitel and Europolitan, Comviq reduced costs by relying on its own
microwave link and introduced consumer-oriented tariff structures after 1994 (Garrard
1998: 268). It also continued to offer a variety of specialised extra services to subscribers
(Holst 1997: 336). In addition, Comvik utilised alternative distribution channels and
sales strategies. ”The company began looking at places where consumers did their
shopping, such as shopping malls, for radio and TV. Since then, Comviq has paid large
commissions ... in order to improve sales” (Mölleryd 1997: 45 - 46).

Generally, this strategy led to positive results: by the end of 1995, Comviq had captured
over 40% of the Swedish market, with two thirds of its subscriber base consisting of
consumers (Garrard 1998: 268). It was only in 1996, however, when the rate of market
penetration in Sweden reached 29%, that Comviq, along with other Swedish network
operators, applied for licensing and was subsequently permitted to implement a high-
capacity DCS1800 network after its existing network reached full capacity (ibid.: 268).
In the Swedish case, notably, implementation of the ‘mass market’ variant of GSM had
followed, rather than preceded, development of the mass market itself. However the
nature of the DCS1800 licenses granted to both new entrants such as Comviq and the
incumbent operator, Telia Mobitel, did support at least one technological innovation.
These were ‘dual band’ GSM licenses, for both 900 and 1800 MHz frequencies. They
therefore created a demand for Ericsson ‘dual band’ hand-sets.

It is possible that Telia Mobitel’s ‘lead’ in implementing a DCS1800 network contributed
to at least a temporary loss of competitive advantage for Comvik. By 1997, Comviq’s
share of the Swedish GSM market had declined from the 40% high-point reached in
1995 to merely 23% (Holst 1999: 114).13 At the same time, however, Sweden’s market
for mobile telephony had grown to a much greater (at least double) size compared to
1994, and Comviq had continued to experience steady growth in the size of its subscriber
base (ibid.: 112 - 113).

5.2.5. Europolitan

Europolitan, which was a comletely new entrant to mobile telecommunications network
operation in Sweden, benefited greatly from the pioneering efforts of Comviq.
Europolitan was established in 1989 as NordicTel, with financing obtained from several

  The measure of of market share here is based on the revenue of operators, rather than the number of

large Swedish corporations. Subsequently, the ownership structure changed, with a 51 %
share of the company being purchased by US-based Air Touch in 1993, when the British
mobile operator, Vodafone, also purchased a 20% share of the company (Holst 1999:
380). 1994 witnessed a public share offering and the sale of remaining shares by several
of the original Swedish owners (Mölleryd 1997: 47).

In 1992, the company, re-named Europolitan, established its GSM network, after a long
and arduous campaign to gain entry to the Swedish market in the face of opposition from
both Televerket Radio/Telia Mobitel and Comviq. Among other things,
NordicTel/Europolitan had to argue against unfavourable initial rulings regarding its
original applications for frequency allocation to the Frequency Management authority,
which was then still part of Televerket (Karlsson 1998: 275 - 277).

Europolitan’s initial marketing strategy resembled Telia Mobitel’s more closely than
Comviq’s: that is, the primary target was the business market, rather than consumers.
Possibly for this reason, Europolitan got off to a weak start in the Swedish GSM market.
However, it quickly followed the lead of Comviq with respect to marketing aimed at
increasing sales to consumers. A variety of tariff structures and the use of retail
distribution channels, as well as agreements with wholesalers, were used to expand
Europolitan’s share of the market. In 1994, Europolitan also launched its own chain of
retail outlets in order to expand sales. (Mölleryd 1997: 47 - 48) In 1996, it obtained,
along with Comviq, a ‘dual band’ GSM license, permitting the company to develop a
DCS1800 network when its 900MHz GSM network had reached full capacity.

Europolitan also continued its original ‘high quality’ marketing strategy, and was the first
GSM operator to obtain ISO-9001 certification for the construction and management of
its mobile networks (Holst 1997: 340). Such measures contributed to a steady growth of
sales over time, and by 1997 Europolitan held a 23% share of the Swedish GSM market,
a position equal to that of Comvik (Holst 1999: 114, 380).14

5.3. Other Private-Sector Actors

As the foregoing account (in section 5.3.3, above) of the main private sector actors in the
development of GSM in Sweden has indicated, the development of new marketing
strategies and distribution channels became increasingly important during the 1990s.
These were means by which the competing network operators sought to expand both the
GSM market as a whole and their respective shares of it.

Businesses that specialised in these areas, particularly in distribution, were thus
instrumental in ‘growing’ the GSM market. Two that have received special mention in
Swedish research are GEAB and Talkline (Mölleryd 1997: 51-52).

  The measure of of market share here is based on the revenue of operators, rather than the number of

5.3.1. GEAB

During the mid-1980s, GEAB was established as a retail shop that sold mobile telephones
and mobile telephone subscriptions. The first store, opened in Stockholm in 1985, soon
became one of the primary sales outlets in Sweden, selling more than 1,000 new
subscribers every year (Mölleryd 1997: 38). During the early 1990s, the business
expanded into a chain of stores whose business was sought after by competing Swedish
GSM network operators and equipment suppliers. GEAB became a major distributor for
Comvik, as well as continuing with sales for Telia Mobitel, but could not reach any
agreement with Europolitan. In 1994, GEAB was purchased as a vehicle for marketing
mobile telephone services across Europe by Unisource Mobile, a subsidiary of Telia’s
international branch (Mölleryd 1997: 51).

5.3.2. Talkline

Talkline, a British mobile telephone company with retail outlets, established a Swedish
store in 1990. Initially, the business floundered, but it later prospered under the
management of a new Swedish proprietor. After the change of ownership in 1991,
Talkline focused primarily on the business market, for which it devised productivity-
enhancing ‘mobile solutions’ that included, among other things, paging services. The
business expanded rapidly, acquiring contracts with a number of large businesses. It sold
subscriptions for all of the major GSM network operators in Sweden, and telephones for
the major suppliers, later enlarging its business to include computers. Talkline
experienced rapid expansion after 1993-1994 and merged with another, more consumer-
oriented, business in 1995 to form one of the leading retailers of mobile phones and
personal computers in Sweden (Mölleryd 1997: 52).

6. Main Outcomes

This section deals with the main economic outcomes of the development of the GSM
standard. There is a primary focus on the growth of the GSM market in Sweden, and an
account is given of factors contributing this development. However, the European and
broader international context is also taken into consideration at the outset. Hence, the
discussion begins with an overall assessment of the development of the GSM standard as
a European ‘success story’. Subsequently, a more detailed account of developments in
Sweden is outlined. Where appropriate, comparisons are made with other nations.

6.1. Consequences: European Competitive Advantage; Nordic Dominance

By the mid-90s, it was already evident that GSM had become a ‘global’ success. In
1996, five years after its launch in 1991, GSM had more than 21 million subscribers in
133 networks operating in more than 105 countries, with 50,000 subscribers signing on
each day (America's Network 1996). This constituted a very substantial share of the total
world market for mobile telephony. Including all known standards, this market was
estimated to amount, at the end of 1995, to some 85 million subscriptions to mobile
telephone services (Holst 1997: 38). The rapid growth of GSM could be partly explained
by the fact that the standard was already well established in Asia and Europe, which
together accounted for 45 million of the total 85 million subscriptions. These regions had
also experienced the fastest growth in the number of subscribers. During 1995, the
number of Asian subscriptions had more than doubled, to 22 million subscribers. In
Europe, growth was less dramatic, but there was an increase of nine million subscribers
in 1995, raising the total number of subscriptions to 23 million. (Ibid.)

By the mid-90s GSM was also recognised as having developed a wider potential market
than any other existing standard in mobile telecommunication -- despite the fact that it
still had little purchase on the largest single ‘regional’ market for mobile telephony, the
US Market. In early 1995, GSM was already, according to one set of estimates,
potentially available to 4.34 billion persons in 94 countries (Holst 1997: 38). The only
comparable coverage by any other standard was that of the ‘first generation’ AMPS, the
US-based analogue standard. In early 1995, AMPS was available to 3.01 billion people in
73 countries worldwide (ibid.). As shown earlier (in section 3.3.), GSM clearly
dominated the ‘second-generation’ market. In 1997, GSM’s only true technological rival,
the US-based IS-95 or CDMA standard (see sub-section 3.2.2.), had between 5.5 and 5.6
million users worldwide. In contrast, the GSM MoU could claim between 4.5 million
and 50 million subscribers. (Shankar 1997: 42)

The success of the GSM standard provided the EU with a singular and much-needed
success in the ICT (information and communication technologies) sector. As one recent
study of Europe and the information and communication technologies ‘revolution’
observes, “The ETSI/GSM project has been of obvious importance for the European
presence in the equipment industry for mobile telecommunications” (Dalum et al. 1999).

This achievement was all the more remarkable and significant against the background of
a general pattern of US and Japanese dominance in the ICT sectors (ibid.). Hence, the
GSM success story – and more generally Europe’s strong performance in
telecommunications – was also a notable exception from which strategic lessons had to
be drawn for the development of future competitive advantage in ICTs.

In addition to these consequences for longer-term competitive advantage, the
development of GSM also had positive economic impacts of a more immediate character.
According to a 1998 report to the European Council, telecommunications markets in the
EU increased in size by one-third and in value by 38 billion ECU between 1995 and 1998
(European Commission 1998: 4). The increase in demand for mobile
telecommunications was identified as one of the main factors contributing to this
spectacular ‘double digit’ growth rate. Based on existing information about market
growth, it was further projected that by 2001, the EU market for mobile
telecommunications would grow to 160 million subscribers and account for the creation
of 400,000 jobs. (Ibid.) It was also stated in this connection that “the GSM market could
more than double to above 170 million customers, if the rest of the EU caught up to
Finland’s current mobile phone density of 50%”, and that this increase in the number of
subscriptions “would mean at least 150,000 new jobs” (ibid).

Despite such optimistic forecasts on the part of some observers, others were quick to
point to the continuing existence of large disparities in subscriber penetration rates and
other aspects of the diffusion of GSM (and mobile telecommunications more generally)
within the EU. According to 1997 statistics from the OECD, Finland was clearly an
exception in Europe and the only countries that even began to approach its remarkable
subscriber penetration rate of 45.6 mobile subscribers per 100 inhabitants were the other
Nordic countries and Italy (OECD 1999a: 76). Sweden had a subscriber penetration rate
of 35.8, Norway had a rate of 38.4, Denmark one of 27.5, and Iceland one of 24. All
other EU countries, with the exception of Italy, at 20.1, had subscriber penetration rates
of less than 20, and some of the largest EU economies, such as France and Germany,
were well below the OECD average rate of 15. (Ibid.) Thus, although the Nordic
countries were undisputed leaders in the world mobile telecommunications subscriber
penetration ‘league’, it was very questionable whether the EU as a whole would ever
catch up with Finland in terms of mobile phone density.

The existence of a large and rapidly growing market for GSM services in the Nordic
countries was matched by the strong performance of Nordic producers of
telecommunications equipment, who – as shown in sub-section 5.1.1 -- became the
leading suppliers of equipment for the GSM standard. By the same token they also
became world leaders in mobile telecommunications equipment more generally.
According to one recent study, two of the three dominant firms in mobile
telecommunications equipment during the 1990s have been Nordic -- Ericsson (Sweden)
& Nokia (Finland) (Dalum et al. 1999). The closest competitor for market share has been
Motorola (US), closely followed by Siemens (Germany). Japanese multinationals have
formed the third rank of competitors. (Ibid.)

Preceding sections have adduced a number of reasons for the ‘Nordic Dominance’ in
mobile telecommunications. Most of them are related to the success of the GSM standard
with whose development Ericsson and Nokia, as well as other Nordic actors in the public
sector, were closely involved. These reasons concern GSM’s relatively rapid and
thorough consolidation of a large regional ‘home’ market as a consequence of concerted
action on standard-setting within CEPT and ETSI, complemented by the market-
expanding liberalisation policies of the EC. (On these points, see especially section 4.)

On the ‘private sector’ side of the ledger, however, there are additional and more specific
reasons for the Nordic producers’ pre-eminence in mobile telecommunications. These
included, naturally enough, the historical tradition of close collaboration between the
Nordic producers of telecommunications equipment and the Nordic PTOs or PTTs, as
well as growing reliance on the output of public research organisations, including
increased hiring of graduates in scientific and technical fields. Firm strategy was also
crucially important, not only with respect to the re-focusing of corporate activity and the
building-up of internal competence, but also with respect to external collaboration, which
was also essential to acquiring key knowledge assets. Arguably, the cross-licensing
agreements between the Nordic producers and U.S.-based Motorola were critical
determinants of the selection of the Nordic prototype technology for the GSM standard
and the subsequent success of Nordic producers such as Ericsson. (On these points, see
especially section 5.)

The formation of a ‘virtuous circle’ in mobile telecommunications in Sweden and other
Nordic countries would not, of course, have been possible without the rapid growth of a
large home market. As other observers have commented, the Nordic experience with
mobile telecommunications provides an instructive example of this phenomenon. “When
a large enough initial market is created, learning effects on the production side may well
provide producers with a great advantage and enable them to develop strong
technological and market competencies which are a solid base for exporting” (Dalum et
al. 1999). The growth of a ‘home market’ for GSM in Sweden depended, as already
noted, on the liberalisation policies implemented by the EC and corresponding public
policy actions in Sweden. However, private sector actors also played important roles in
‘growing’ the market, not only through competition among network operators and service
providers, but also through entrepreneurial marketing of services and equipment. (On
these points, see especially section 5.) Some of the main consequences for growth will
be examined more closely towards the end of this section.

6.2. Changes to Charging Systems

Before presenting statistics on the growth of mobile telecommunications equipment an
services in Sweden, it will be useful to examine further factors contributing to the growth
of the market for GSM and other mobile telecommunications services. In section 6.1,
these were discussed in a rather general way, in terms of ‘liberalisation’ and
competition’. Here, further insight can be developed by focussing attention more closely
on how liberalisation and competition were manifested in changes to charging systems.

In section 3.3, some of the main reasons for the relatively poor international performance
of U.S.-based standards for ‘second generation’ mobile telecommunications were
summarised. The discussion referred not only to a ‘divided market’ but also to the weaker
migration in North America than elsewhere from ‘first generation’ to ‘second generation’
standards. Both were due to regulatory decisions that stressed the necessity of achieving
‘backwards compatibility’ with the pre-existing analogue standard, AMPS, rather than
unified or inter-compatible digital standards. It can also be observed here that decisions
regarding charging were another factor contributing to the relatively slow growth and low
subscriber penetration rates for mobile telecommunications within the US.

According to recent statistics from the OECD, the US market for mobile
telecommunications had a subscriber penetration rate in 1997 of just over 20% (OECD
1999a: 76). This is far below the rates cited in section 6.1 for the Nordic countries.
Moreover, there had been a relatively slow market growth up to that point, characterised
by “steady progression” and a continuing dominance of the analogue AMPS standard,
such that “the introduction of digital services was … a gradual transition that had an
almost imperceptible effect” (Garrard 1998: 348). The absence of rapid growth
accompanying the introduction of digital standards in the US may be explained by the
existence of an adequate infrastructure based on the successful analogue standard,
AMPS. But explanations for the more generally sluggish growth of mobile
telecommunications in the U.S. must be sought elsewhere. One important factor of
explanation was clearly the structure of charging, or tariffs, on mobile services.

There were clear disincentives to subscribe to mobile services in the US, since, in
addition to paying fixed subscription rates, (varying) call charges and fees for extra
services, such as roaming, “cellular users in the US were also required to pay for
incoming calls” (ibid.: 44). The utility of mobile services was greatly reduced, since
users were understandably anxious to avoid ‘junk calls’ and relucant to give out their
mobile numbers to people they did not know very well. Although operators and service
providers attempted to overcome this obstacle by developing a variety of tariff packages,
the problem persisted well into the 1990s, when the Clinton administration introduced a
new telecommunications act. Among other things, the new legislation enabled cellular
operators “to offer long-distance services to their customers along with their cellular
packages, instead of handing local calls over to one of the long-distance (Inter-Exchange)
carriers” without having to provide equal access to them (ibid.: 354). These changes
made it possible to simplify the structure of charges for interconnection, thereby lifting
the main barrier that had previously prevented the introduction of “caller-pays” charging.

Although these were rather late developments for the US market, similar measures were
introduced at a relatively early point in Europe, with positive effects on the increase of
subscriptions, at least so far as the market development of GSM was concerned. As
discussed in sub-section 4.2.3, the GSM Memorandum of Understanding (MoU) that was
first developed in the late 1980s at the urging of the EC had required the speedy
resolution of such ‘commercial issues’ by the signatories. In that context, ‘roaming’ and
interconnection fees had proved, as in the US, to be a difficult issue. However, it was

resolved much earlier and more decisively in favour of implementing a ‘caller pays’
system, with additional costs for international roaming assigned to the recipients of calls.

As discussed in sub-section 3.1.2, with reference to the introduction PCN (personal
communications networks) standard in the UK and its subsequent adoption by ETSI in
the form of the DCS1800 standard, consumer-oriented standards based on high-density
urban networks were essential to the rapid market growth of mobile telecommunications
in Europe. As also stressed in that discussion, the success of these ‘mass-market’
adaptations of GSM in the UK depended critically upon competitive tariff structures.
There were essentially no major technical between GSM and PCN/DCS1800, but
terminal equipment for the latter had to be sold at lower prices, and the services provided
for it required much higher infrastructure costs. These conditions made it imperative for
PCN/DCS1800 network operators and service providers to resort to purely ‘commercial’
means of achieving the rapid and extensive market growth that would be necessary in
order to cover their large initial investments in infrastructure and equipment. In addition
to aggressive marketing and the development of new distribution channels, such means
included the use of alternative tariff structures and the re-negotiation of interconnection
fees with fixed networks (PSTNs) in order to lower the price of subscriptions and charges
for network use. The sacrifices of network operators eventually made possible an
expansion of the UK market to a subscriber penetration rate of 12% by 1996.

Sweden, as pointed out in sub-section 5.2.3 exhibited a different pattern than the UK with
respect to the diffusion of GSM, including its mass-market variant, DCS1800. There
were several factors contributing to the markedly more rapid and higher rate of market
growth in Sweden. First of all, the initial rate of market penetration already achieved by
mobile telecommunications during the ‘first generation’ of analogue standards was much
higher in Sweden – at least three times that of the 1990 UK rate of 2%. This higher
density was largely due to the beneficial effects on market development of the earlier
selection of a common analogue standard, NMT, by the Nordic PTOs and PTTs. Second,
Sweden’s basic subscription rate had been kept to about half that of the UK, even though
there had been a relative lack of competition within the Swedish market. While the lower
rates in Sweden may have been due to the ‘public service orientation’ of
Televerket/Telia, they could also be readily understood as sacrifices providing the means
of expanding the market quickly to a size where it could ‘naturally’ sustain lower pricing.
Third, increased competition resulting from ‘liberalisation’ of the Swedish market did not
reduce already low prices for subscriptions. It only helped re-orient service marketing
strategies more towards consumers and non-business users. For all these reasons, the
introduction of DCS1800 services in Sweden did not mark the advent of a ‘mass market’
for mobile telecommunications, pioneered by ‘new entrants’. This development was
already well under way due to standardisation and market development strategies already
initiated in connection with the NMT standard during the ‘first generation’ of analogue
mobile telecommunications. Moreover, the DCS1800 market was ‘pioneered’ by the
incumbent, Telia, and not the new entrants. The main innovations of the latter lay in the
areas of marketing strategies, such as the introduction of pre-paid card subscriptions.

The far-sighted pricing policies of Nordic operators and service providers in the early
1990s, thanks to their positive effects on market growth, appear to have been fully
vindicated by the late 1990s. At that time it became possible to introduce new strategies
for further expansion of the market that were entirely different from varied tariff
structures aimed at capturing niche markets. Hence, the 1999 issue of the OECD’s
Telecommunications Outlook makes the following observations on recent and future
developments in the pricing of mobile telecommunications:

              As with most competitive segments of the fixed market, competition is
              obliging operators to simplify tariff structures for consumers. The
              Scandinavian mobile communication market is often a good indicator of what
              will happen in other OECD countries. In this regard, the move by Telia to
              reduce the number of mobile pricing options from ten subscription plans to
              two is noteworthy. This initiative invites comparison with the more
              simplified rate structures in the fixed network for the long-distance market,
              brought about by maturing competition in the United States. In the future, it
              would be a welcome development if the same impact was to occur for calls
              between fixed and mobile networks in terms of international roaming.
              (OECD 1999a: 76).

6.3. Growth Statistics15

Against the background sketched in other parts of this section (the preceding sub-
sections), it is now possible to present statistics concerning the growth of markets for
telecommunications equipment and services in Sweden during the ‘first’ and ‘second’
generations of mobile telecommunications. The preceding discussion should be referred
to for explanation and interpretation of the information presented in the sub-sections

6.3.1. Growth Figures (Equipment)

In the graph below the number of terminals sold in Sweden during the period 1982 to
1999 is shown. (See figure 6.1.)

  The primary author of this sub-section is Esa Manninen, who contributed not only the lion’s share of the
text but also all of the statistical data presented in the form of figures and tables.

                         Figure 6.1: Number of terminals sold in Sweden
                                      (thousands), 1982-1999

                         300                                                           2000
  Number of NMT450 and

  NMT900 termnals sold


                                                                                              Number of GSM
                                                                                               terminals sold

                         150                                                           1000
                          0                                                            0
                               82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99

                                         NMT450          NMT900          GSM

Note that the number of GSM terminals sold follows the scaling on the right-hand Y-axis.
Source:       MobilTeleBranschen, 1999 and 2001.

It is clear that each generation of terminals have been sold in ever increasing quantities.
The sales of NMT450 terminals peaked at about 41 000 units sold in 1988. In 1994 sales
of NMT900 terminals reached its maximum of 274 000 units sold. The GSM terminals
were introduced on the Swedish market in 1992, and the sales of these terminals is still

The numbers of GSM terminals sold in Sweden each year are given in the following

Table 6.1: GSM Terminals Sold in Sweden Each Year, 1992 – 1999.

Year                         1992 1993 1994 1995 1996 1997 1998 1999
Number of sold terminals     3     44      367 764 925 1203 1546 1746
(in thousands)
Source: MobilTeleBranschen, 1999 and 2001.

The growth in number of sold GSM terminals has been quite dramatic in Sweden since
the market introduction in 1992. It was mentioned earlier that the sales of NMT900
terminals peaked at 274 000 units sold in 1994. During the same year about 367 000 units
of GSM terminals were sold. It took the NMT900 terminals 8 years on the market to
reach the maximum number of units sold. The GSM terminals, on the other hand, passed
the NMT900 maximum during the third year after market introduction.

6.3.1. Growth Figures (Services)

During the year 1994 the growth in GSM subscribers for the first time exceeded the
growth in NMT900 subscribers in Sweden (Lernevall and Åkesson, 1997: 571). In mid-
1995 the distribution of GSM subscribers between the three actors in Sweden were: 355
000 for Televerket, 307 000 for Comviq, and 93 000 for Europolitan (ibid.: 571).

One important reason for this rapidly increasing number of subscribers was the mobile
telephone operators’ heavy subsidisation of the subscriptions (ibid.: 571). For the
consumer it meant that a new cellular terminal could be bought for a symbolic price.

In the following table (6.2) a more detailed picture of the change of number of
subscriptions in the Swedish GSM system is given. Note that the figures for Comviq are
reported under the heading Tele2. At the turn of 1997/98 Comviq GSM AB and Tele2
AB merged, and they are now owned by NetCom Systems AB (Post och Telestyrelsen
1998: 2, n. 1; PwC and Öhrlings-Price-Waterhouse-Coopers 1999: 32).

Table 6.2: Total Number of GSM Subscriptions during the Period 1994-1998 in
Sweden, and the Shares for Each Mobile Operator.16

YEAR                          1994 1995                  1996      1997      1998
Total number of subscriptions 422    1 033               1 571     2 414     3 605
(in thousands)
Telia                         51% 45%                    52%       49%       47%
Tele2                         32% 41%                    30%       34%       36%
Europolitan                   17% 14%                    18%       18%       17%
Source: PwC, 1999: Appendix B, table 5.

The most striking feature of the above table (6.2) is the rapid increase in the total number
of subscriptions during the period. An increase from below 0.5 millions to 3.6 millions in
four years only. The highest growth rate was for the period 94-95, when the number of
subscriptions increased by 145%. For the remaining years the yearly growth rate was
about 50%. During this period of high total market growth the largest Swedish operator
Telia has managed to keep about half of the market in terms of number of subscriptions.

Pre-paid card mobile telephony was first introduced by Tele2 in the spring of 1997. Later
the same year Europolitan introduced their card, and they were followed by Telia in July,
1998 (PwC, 1999: 42). The table (6.2) above showed the number of GSM subscriptions

  Apparent differences between the figures presented in this table and information on the market shares of
the major Swedish operators presente earlier, in sub-sections 5.2.3 – 5.2.5, can be explained by the fact
that this table refers to subscriptions, whereas the previously quoted figures referred to revenues.

including pre-paid cards. In the next table (6.3) the pre-paid cards subscriptions are
separated from the other subscriptions.

Table 6.3: Pre-Paid Card Subscriptions in Sweden, 1996 - 1998

YEAR                                    1996 1997 1998
Number of pre-paid cards subscriptions -       235      1 023
(in thousands)
Other forms of subscriptions            1 571 2 179 2 582
(in thousands)
Total number of subscriptions           1 571 2 414 3 605
(in thousands)
Source: derived from figures in PwC, 1999: Appendix B, table 5.

It can be seen that the growth rate has been much higher for the pre-paid cards than for
the other forms of subscriptions between 1997 and 1998. While the other forms of
subscriptions increased by 18% during the period, the growth rate for pre-paid cards
subscriptions was 335%.

Growth in the mobile telephony services production could also be described in terms of
revenues of the operators. In the next table (6.4) the total value (in terms of revenues) of
the Swedish mobile services market is separated into the NMT and the GSM segments.

Table 6.4: Revenues of the Swedish Mobile Services During the Period 1994-1998, in

Year            1994 1995 1996              1997    1998
Total revenues 4 340 6 050 7 420            8 420   11 349
GSM revenues 1 070 2 310 4 460              6 190   9 883
NMT revenues 3 270 3 740 2 960              2 230   1 466
Source: PwC, 1999: Appendix B, table 6.

The source for table 6.4 does not indicate whether or not these are nominal prices.
However, there was very little inflation during the period covered by this table.

From the above table (6.4) we can see that the total revenues almost doubled between
1994 and 1997, and in 1998 the total revenues exceeded 11 billion SEK. But it should be
noted that in the above table the operators’ revenues from interconnection fees are

excluded.17 The table also reveals that revenues from the GSM system exceeded the
revenues from NMT for the first time in 1996.

   Interconnection fees are paid by operators to get access to other operators’ (fixed and/or mobile)
telecommunications networks. For 1998 it was estimated that the interconnection fees made up something
between 20-30% of the operators total revenues. Including interconnection fees, the estimated total
revenues in 1998 was 15.1 billion SEK (PwC, 1999: 30; Appendix A).

7. Conclusions

As explained at the outset, this report is only one part of a larger ‘work-in-progress’.
Hence, any ‘findings and conclusions’ developed here can have only a very limited scope
for generalisation. Nevertheless, it may still be useful to develop some preliminary
statements that can be applied within the framework of the larger research project.

The main point of departure for developing preliminary findings and conclusions is
provided by a series of ‘key questions’ that have been elaborated in relation to this and
other projects within the research network on European Sectoral Systems of Innovation
(ESSY). These questions, listed below, can be used to assess the empirical material
assembled in this report.

To begin, then, it will be appropriate to re-state the ‘key questions’ of the ESSY research
network. According to internal documents generated by the network, these questions
concern the following topics:

              - Knowledge Base & Learning Processes
              - Firms, Non-Firm Organisations & Networks
              - Geographical Boundaries
              - Long-Term Dynamics of the Sector and Co-Evolutionary Processes
              - Public Policy
              - European International Performance and Comparisons with US & Japan

Below, preliminary findings and conclusions will be developed in relation to each of
ESSY’s key questions. The various topics are addressed in the order that they appear

7.1. Knowledge Base & Learning Processes

This report has examined the evolution of the mobile telephony during the second of
three successive ‘generations’ of technical development. It has observed that the
knowledge-base of mobile telephony became increasingly more complex and diffuse
during that period. This transformation was reflected in the changing division of labour
between the public and private sector actors in mobile telephony.

In the first generation of mobile telephony, based on the convergence of radio and
telephone technologies in telecommunications, most of the relevant knowledge base was
internalised -- i.e., contained ‘in house’ -- by a number of discrete organisational actors in
both the public and private sectors. For the most part, the private sector actors possessed
specialised expertise in various subsystems -- for example, switching, radio systems, etc.
In contrast, the main public sector actors -- i.e., the PTOs or PTTS -- possessed what
could be called ‘system competence’. In Nordic countries such as Sweden (the PTOs or

PTTs) acted as ‘system integrators’, and this capability was reflected in the leading role
that they played in defining the first-generation NMT standard.

In the second-generation of mobile telephony, the same division of labour was at first
maintained, but then began to erode. This was due to a number of factors, related to
institutions, technologies and markets. Institutional factors were, perhaps, most visible.
In Europe, regulatory reforms by the EU increasingly confined the role of the PTOs or
PTTs to that of network operation. Separate regulatory authorities were established, and
the PTOs or PTTs lost their monopolies over service provision. Standards-making, once
their exclusive domain, was opened up to a much broader range of actors, including
suppliers and users.

These reforms were, however, partly related to increased technological complexity. The
convergence between mobile telephony and information technology created a
discontinuity that necessitated the entry and involvement of other actors in the formerly
‘closed’ field of telecommunications. Moreover, market forces were also important.
Heightened international competition in the telecommunications industry led to greatly
increased expenditures on R&D. In turn, this meant that ‘national’ equipment suppliers
and national PTOs or PTTs could no longer simply rely on one another’s complementary
assets in order to acquire and develop ‘in house’ the knowledge they required to remain

These changes in the knowledge base appear to have been primarily the result of
technological and market developments. As a consequence of alterations in the
knowledge base, the key actors in mobile telephony, both public and private, found it
increasingly necessary to form strategic alliances with a wide range of other actors, often
across national boundaries. Increased interaction with public research organisations also
became necessary. All of these considerations influenced the shift from ‘closed’ to
‘open’ standards development. In this instance, at least, institutional change ‘followed’
and formalised organisational changes caused primarily by technologies and markets,
rather than actually initiating such change. In Sweden and Europe, the technological and
organisationa l changes that characterised the ‘second generation’ of mobile telephony
were not preceded by broad institutional changes such as deregulation. Rather, they
occurred either earlier or at much the same time, thereby providing a technological and
organisational platform for the implementation of such policy measures.

Now that the second-generation of mobile telephony has entered its mature phase, it
appears that the PTOs or PTTs have given up a large part of their former role as systems-
integrators. Even in the early phases of GSM’s development, the ‘system competence’
formerly associated with this role appeared to have already been transferred to major
equipment producers. Televerket’s largely ‘symbolic role and Ericsson’s de facto
leadership in the technical development of the ‘Nordic’ proto-type technology for the
GSM standard demonstrates this point. Moreover, PTOs or PTTs lost their ‘official’
recognition as the only important public sector actors capable of acting at a ‘systems’
level. SDOs such as ETSI assumed increased importance in this respect, not least
because of their capacity to involve many other actors in developing and defining

standards. Public research organisations also became increasingly important to the
further development of the technological knowledge base.

Among private sector actors, the major telecommunications equipment firms remained
most important. For these firms, the build-up of competence through a variety of means
was critical to their success in the ‘second generation’ of mobile telephony. Ericsson, in
order to focus its activity on mobile telecommunications resorted to corporate
restructuring, mergers and acquisitions, joint-ventures and various other kinds of
collaboration, in addition to greatly increased expenditures on R&D and the hiring of
skilled personnel. Thus the firm’s ‘narrowed’ focus on mobile telecommunications was
accompanied by the development of a much broader and more internally diversified
profile of competence. Generally, telecommunications equipment producers appeared to
pursue two different but complementary strategies simultaneously. On the one hand, they
developed their own capabilities as system integrators. On the other hand, they entered
into a greater number and wider variety of ‘strategic alliances’ with other actors. In the
case of Ericsson, which exemplified both of these strategic trends, its cross-licensing
agreements with Motorola were the key to its development of a complete ‘system
competence’ with respect to GSM.

7.2. Firms, Non-Firm Organisations & Networks

The organisational developments sketched or alluded to above include the transition from
closed to open standards making, which led to the emergence of a qualitatively different
type of SDO (e.g., the hand-over from from CEPT to ETSI, and corresponding changes in
other regions). There has been an accompanying shift in terms of ‘key actors’ from
multi-purpose and monopolistic organisations (the PTOs/PTTs and major equipment
producers) to a new division of labour in which ‘key’ roles are more widely distributed
among a greater variety of actors. Apart from the SDOs, which represent a multiplicity
of only partly coinciding organisational interests, there are no longer any true
monopolists in the global telecommunications sector, much less the world market for
mobile telephony. Even the largest discrete organisations (the PTOs/PTTs and major
equipment producers) are increasingly performing specialised roles, although they retain
and seek to develop certain capacities for system-integration. On this basis, it can be said
that the evolution of mobile telephony has involved progressively greater functional
differentiation and complexity in the division of labour.

The increasing variety of non-firm organisations involved in mobile telephony has
already been commented on. The shedding of ‘system competencies’ by the PTOs and
PTTs was accompanied by a shedding of roles and responsibilities, many of which were
assumed by other public sector organisations. Special attention been paid to 1) the
increasing importance of SDOs and 2) the greater differentiation of roles and finer
division of labour amongst public-sector actors. The increased diversity of non-firm
organisations in the public sector created a need for new mechanisms of co-ordination
that was only partly met by new SDOs such as ETSI. Governments -- in Europe, the EC,

in particular – became increasingly involved utilising in telecommunications standards as
vehicles for industrial and trade policy.

With respect to specific organisational forms, such as firms and networks, it appears that
what has occurred in the definition of ‘second generation’ standards in mobile telephony
has been the formalisation and consolidation of networks that were previously more
informal in character. Among both non-firm organisations and firms for example,
formalisation took the form of establishing consortia or strategic alliances, based on
agreements such as the GSM Memorandum of Understanding (MoU). Consolidation, at
least amongst firms, often took the form of mergers and acquisitions, such as Ericsson’s
nearly complete absorption of the entire telecommunications equipment sector in

In comparison to the first generation of mobile telephony, moreover, these developments
occurred at a higher level of integration in the second generation. The NMT standard
emerged within one region in Europe, but the GSM standard was developed within
Europe as one region in the world. In that context, the creation of the GSM MoU
provided a vehicle for overseas expansion of a market that had originally been defined on
a regional basis. Formalisation thus constituted a basis for further of expansion, which in
turn entailed the formation of new network relationships as GSM became effectively a
‘world standard’. Formalisation, moreover, led to the development of new co-ordination
mechanisms, or the redefinition of relations among existing ones. At critical junctures,
new forms of governance emerged as means of consolidating the new order. The
creation of ETSI, representing a new regime of ‘open’ standards making, is a case in

The development of new kinds of governing bodies within the public sector was
accompanied by the proliferation among private-sector actors of partly over-lapping and
partly opposing consortia, coalitions, alliances, etc. (Some of these will be discussed
below, in sub-section 7.3 on ‘Geographical Boundaries’.) This observation raises the
question of whether or not an expanded taxonomy of organisational forms and modes of
governance might be necessary in order to develop an adequate organisational and
institutional ‘map’ of mobile telephony. One example of such a taxonomy includes
markets, hierarchies, states, networks, and associations as ”distinctive modes of
governance” (Hollingsworth, Schmitter, and Streeck 1994: 8). These different
arrangements for co-ordinating economic activity are all likely to be present in varying
degrees, and articulated with one another in different ways, in any given economic sector.

7.3. Geographical Boundaries

With respect to geographical boundaries, it is evident that standard-setting in second-
generation mobile telephony has involved a higher level of internationalisation than in the
first generation. The NMT standard was conceived primarily as a regional standard,
though it later verged on becoming pan-European. Building upon this experience, most
second-generation standards were developed as regional standards with the potential to

become de facto world standards through international adoption. (The decision to make
second generation US standards ‘backwards compatible’ with the pre-existing analogue
standard, AMPS, thus capitalising on the latter’s wide market distribution, could be
interpreted as a variant of this strategy.) GSM more than fulfilled the expectation of wide
international diffusion. Initially conceived as a pan-European standard, it gathered
sufficient momentum to become the leading ‘world’ of its generation.

Despite greater internationalisation, open competition with foreign standards never
occurred in the regional (national or multi-national) ‘home markets’ with which the main
second generation standards in mobile telecommunications were associated. In these key
markets, co-operation between standards development organisations, standard-setting
agencies and regulatory authorities resulted in decisions that effectively partitioned the
international market. Japan, perhaps the most extreme case of a closed market, generated
a ‘closed’ standard with little potential for international diffusion. The US market, which
allowed only internal competition, was divided between two domestic standards. As a
result of their rivalry in the US, neither managed to dominate the ‘open’ portion of the
international market. GSM also developed within a sheltered regional market, but one in
which, unlike the US, the standard-setting and regulatory regime did not require
backwards-compatibility with a well established ‘first generation’ standard. Within the
boundaries of this market, in which no first-generation standard had truly dominated, a
particulary rapid migration from first- to second-generation standards occurred, providing
GSM with sufficient momentum to capture a large share of external markets.

The sustained development of a strong and constantly expanding ‘home market’ for
mobile telecommunications was clearly one of the factors that contributed to the
competitive advantage of equipment producing firms who had been involved with the
development of the standard(s) defining that market. Thus, the dynamism of the EC
market for GSM helped European firms such as Ericsson and Nokia to become leading
producers of terminal equipment and infrastructure for mobile telecommunications
world-wide. These developments are often attributed to the market-expanding effects of
the liberalisation policies ushered in by the EC during the 1990s. However, closer
inspection of the sub-regional Nordic market that was ‘home’ to Ericsson and Nokia
yields a somewhat different picture.

While the EC’s liberalisation policies have had a very uneven effect on the diffusion of
mobile telecommunications across the European Community, the Nordic countries had
achieved some of the world’s highest rates of subscriber penetration for mobile
telecommunications even before the advent of liberalisation. This was largely due to the
consolidation of a strong market for mobile telecommunications via concerted action by
the Nordic PTOs and PTTs in defining the first-generation NMT standard. In Sweden,
where a competitive market only began to develop at a rather late point, Televerket/Telia
kept prices for subscriptions to second generation GSM services very low throughout the
1990s, effectively subsidising GSM subscriptions in order to achieve rapid and extensive
market growth. Moreover, it was Televerket/Telia, rather than the new entrants, that
pioneered DCS1800, the ‘mass market’ variant of GSM. For these reasons, it can be
argued that, at least in the Nordic sub-region, enlightened standardisation and pricing

policies on the part of powerful incumbents contributed at least as much as liberalisation
to the development of the GSM ‘home market’.

A similar strategy for creating a stronger home market seems to be implied in recent
recommendations by Dalum and Villumsen (2001) on how to strenthen the competitive
position of European producers of equipment for fixed (and mobile data communication).
They argue that “one of the most relevant policy instruments at hand appears to be to
create massive incentives for increasing the penetration ratios of broadband access among
consumers in Europe” (ibid.: 16).

Given its beneficial effects on the development of producer firms’ technological and
market competence and their capability to face competition on export markets, ‘learning’
from a dynamic home market should not be discounted as an important source of
competitive advantage. However, such learning was not the only source of the
technological and market leadership in mobile telecommunications that Ericsson and
Nokia acquired via their involvement with the development of the GSM standard. There
were additional influences and mechanisms at work and not all of them involved close
geographical proximity, although geographical boundaries shaped their interaction.
Especially important in this connection were collaborative relationships and strategic
alliances between firms based in different regional markets. Both before and after its
success with GSM, Ericsson employed these means to expand into international markets,
internationalising its R&D and becoming a producer for multiple standards.

GSM was from the very outset the project of a regional consortium, eventually
consolidated By ETSI under the tutelage of the European Commission. However, this
consortium was internally divided by both national coalitions and strategic alliances
amongst firms. The national coalitions were generated by the historical relationships of
near vertical integration that had obtained between major equipment producing firms and
the PTOs or PTTs that formed the membership of CEPT and later the ‘first rank’ of
menmbership within ETSI. Contention between these coalitions was a marked feature of
the process by which CEPT selected a prototype technology for the GSM standard. In
that context, a Nordic coalition confronted a Franco-German one, and both were aware
that the outcome of the selection process would confer a strong competitive advantage on
producer firms that had been involved in developing the winning prototype technology.
It was this awareness that prompted US-based Motorola to provide Nordic producers
such as Ericsson and Nokia with access to technology that was crucial to their success in
the selection of a prototype for the GSM standard. Subsequently, Ericsson and Nokia
became leading producers of equipment for the GSM standard. Motorola, although it had
only very limited possibilities to participate directly in the growth of the GSM market,
could still benefit from the revenues generated through its strategic licensing of fully half
of the essential patents for the GSM standard.

The regional consortium that developed the GSM standard was thus shot through with
intersecting national and intra-sectoral (or inter-firm) fault lines that formed the basis for
strategic manoeuvring on the part of important actors. These divisions gave rise to
conflicts whose resolution shaped both the GSM market and positions within it. The

conflict that arose over the procurement policies of the GSM MoU administration is an
instructive case in point. Originally, the GSM MoU sought to appropriate all intellectual
property rights essential to the GSM standard as a condition for the award of
manufacturing contracts to the equipment producing firms. This policy was strongly
resisted by the equipment producers, led by Motorola. Eventually, ETSI resolved this
dispute in accordance with the EC’s recognition that if European Community markets
were to attract and benefit from private investments in R&D, IPRs would have to be
respected and compensated. ETSI’s compromise solution allowed for cross-licensing
among equipment producing firms. Large firms, which had considerable assets in the
form of IPRs were satisfied with this arrangement, but it disadvantaged smaller firms
without comparable resources. In effect, the solution was one that favoured market
concentration and oligopoly and raised steep barriers to market entry. Firms such as
Ericsson and Nokia were able to consolidate commanding positions within the market for
GSM equipment and the possibilities for European-based network operators to award
manufacturing contracts to low-cost producers outside of the EC were greatly reduced.

7.4. Long-Term Dynamics of the Sector and Co-Evolutionary Processes

Some of the main observations that can be made here have already been stated under
other headings. For example, it was observed in relation to the ‘knowledge base and
learning processes’ of mobile telephony that there has been a clear trend, over the three
‘generations’ studied here, for the knowledge-base to become increasingly more complex
and diffuse. As a result, there has emerged a more extensive and detailed division of
labour between public and private sector actors in mobile telecommunications, as well as
within each of these categories. Under the heading of firms, non-firm organisations, and
networks, the observation was made that mobile telecommunications has evolved
towards a progressively greater functional differentiation and complexity in the division
of labour, especially in the public sector. This has been paralleled by the growth of
strategic alliances, coalitions and consortia in the private sector. Under the same
heading, it was also observed that there has been an ongoing dynamic of ‘formalisation’,
since each successive standard in mobile telephony has entailed the formalisation and
consolidation of networks that were previously more informal in character. With respect
to ‘geographical boundaries’, it was remarked that each successive standard has occurred
at a progressively higher level of ‘internationalisation’.

Many aspects of this pattern of development -- for example, the increase in ‘functional
differentiation’ and the trend towards greater and more extensive ‘formalisation’ --
suggest that mobile telecommunications has closely followed the well-known ‘industry
life cycle’ and is now in or approaching a stage of maturity. However, there are also
obvious departures from this pattern. For example, rather than simply becoming more
concentrated and oligopolistic, the equipment industry has continued to witness the
emergence of ‘new entrants’ from other sectors. Similar developments have occurred in
network operation and service provision, where – perhaps more for regulatory than for
technical reasons -- the clear trend is away from monopoly and towards greater variety

and more competition. These developments indicate that processes of technological
renewal, combined with ‘liberalisation’ initiatives, have forestalled ‘maturity’.

In some respects, technological renewal was a matter of increase in ‘degree’, rather than
‘kind’. For example, there was an increasing use of computer software in each
successive generation of mobile telecommunications. The fully digital second-generation
systems required more digital signal processing compared to the first-generation analogue
systems. Consequently firms such as Ericsson began to pose much greater demands on
national education and training systems for an adequate supply of computer science
graduates and other categories of highly skilled labour. In addition, however, the
generational shift in mobile teleciommunications also marked important changes in the
quality or character of the competence requirements of firms. These processes of
technological renewal were based, to a great extent, on the convergence of formerly
separate technologies. In the first generation, telephony and radio were combined. In the
second, digital technology was fully implemented, creating possibilities for data
transmission, in addition to voice transmission. Thus, rather than marking developmental
stages within a single technologica l trajectory, the succeeding generations of mobile
telecommunications represented new technological paradigms. Each opened up a broad
range of evolutionary possibilities for market and institutional forces to respond to.

The implications of this and similar developments have already been discussed at some
length under the heading of ‘knowledge base and learning processes’ and ‘geographical
boundaries’. Producer firms such as Ericsson needed to develop a much broader and
more internally diversified profiles of competence in order to develop second-generation
systems, and the strategies to which they resorted in order to meet these requirements
placed new strains on existing institutional set-ups. The institutional arrangements based
on ‘quasi-vertical integration’ between PTOs/PTTs and equipment producing firms that
CEPT attempted to perpetuate in the initial procurement policies of the GSM MoU
administration failed to withstand these strains, as discussed in sub-section 7.3, on
‘Geographical Boundaries’. As a consequence, important institutional changes in
relations between these two categories of actors were eventually brought about under the
aegis of ETSI and the EC.

The past record of rapid growth in mobile telecommunications has, of course, been the
outcome of both supply and demand forces at work. The interaction between these
market forces has affected generational processes of technological renewal. One
important reason for introducing the GSM standard was to increase the subscriber
capacity. In this respect, technological development on the supply side clearly responded
to a well defined aspect of existing demand. In addition, however, the GSM standard
offered enhanced services like pan-European roaming, and improved voice quality, and
capabilities for data-transmission. All of these technical features of the standard provided
new bases for market development. In these respects, technological development on the
supply side helped to create new forms of market demand. With the creation of ETSI and
a new regime of ‘open’ standards development during the later stages of the GSM
standard’s development, these relationships were officially recognised and
institutionalsed as an essential part of the process of standard development.

Conclusions concerning the ‘knowledge base’ suggested a considerable lag between
technological and institutional developments. For example, ‘digitisation’ is recognised as
the technological basis on which it became possible to separate telecommunications
network operation from service provision. It was therefore possible to make this kind of
separation between ‘infrastructure’ and ‘services’ when the second generation of mobile
telecommunications first appeared. Nevertheless, this separation was not fully
implemented with second-generation standards such as GSM, despite the best efforts of
the European Commission. Incumbents and regulators within national
telecommunications markets were able to resist such attempts to maximise competition in
mobile telecommunications.

Notwithstanding this laggardly pattern of institutional change, the technological evolution
of mobile telecommunications has been, to a remarkable extent, a managed process.
Thus, technological discontinuities in mobile telecommunications that were potentially
competence-destroying for incumbent organisations have been introduced in a manner
that has been, for the most part, competence-enhancing. Institutions and rule-making
authorities -- in particular, the SDOs -- have, moreover, played a leading part in
managing the incorporation of new technologies and new sources of competition. Thus,
there has been a highly regulated co-evolution among markets, technologies and
institutions, which can not be adequately described with simple terms such as
‘institutional lag’ -- or, for that matter, ‘technology push’ and ‘market pull’.

7.5. Public Policy

For some time now, the ‘conventional wisdom’ concerning policies for innovation in the
telecommunications sector (and others) has favoured the liberalisation of markets,
complemented by supply-side initiatives aimed at increasing industrial R&D efforts,
often through increased interaction with public research organisations. Thus,
‘deregulation’ leading to increased competition is often identified as the primary cause of
higher rates of innovation and lower costs for ‘standard’ products such as ‘plain old
telephone service’ (POTS).

However, the development of mobile telephony in the European Community -- which has
arguably enjoyed remarkable success with this particular set of innovations in
telecommunications -- has not been due primarily to market liberalisation and strong
supply side initiatives. The first two generations of mobile telephony were conceived and
introduced primarily by PTOs or PTTs. These actors were essentially public sector
monopolists that also exercised regulatory powers in closed national markets.
Deregulation of the telecommunications sector and strong supply-side programmes in
support of advanced R&D in telecommunications were introduced by the EC only after
the second-generation standard, GSM, had been defined and was well on its way to full
implementation. (See, for example, the discussion of Nordic market development in sub-
section 7.3, under the heading of ‘Geographical Boundaries’). The real test of the new
policies, therefore, will be the success of the third-generation standard, UMTS. In the

meantime, ‘conventional wisdom’ is challenged by the conditions under which the earlier
standards, NMT and GSM, achieved their success.

One lesson that the EC -- and, possibly, other governmental authorities -- have clearly
drawn from experience with the last two generations of mobile telephony is the strategic
importance of standards as a means of creating ‘organised markets’ and co-ordinating
supply and demand. Indeed, recent EU policies regarding telecommunications standards
have been explicitly motivated by considerations of European ‘competitive advantage’.

In creating ETSI and introducing an ‘open’ approach to standards development, the EC
did not simply bow to growing pressure to include an ever wider variety of relevant
actors in the development of telecommunications technology. Similarly, by initiating co-
operative international research on mobile telecommunications, the EC has not merely
acquiesced to an increased fragmentation of the public-sector’s role in the development
of new technologies in telecommunications. Rather, there has been a clear effort to
provide effective co-ordination mechanisms at higher levels of integration -- i.e., the EU
and ‘world’ levels -- than those which were previously addressed in ‘national’ and
‘regional’ policy initiatives.

The policy record of the EC with respect to telecommunications clearly indicates that it
regards standards and standards development organisations as important vehicles for
industrial and trade policy. Of course, the ‘effectiveness’ of these vehicles is another
question, as is that of the related research programmes.

7.6. European International Performance and Comparisons with US & Japan

The ‘conventional wisdom’ concerning Europe’s international performance in mobile
telephony, as compared to that of the U.S. and Japan, is that Europe has emerged as a
clear leader in mobile telephony due to its success in defining standards in mobile
communications. Ericsson’s and Nokia’s dominance among equipment producers in
mobile telephony is often traced to the early success of the NMT standard, and GSM is
similarly regarded as the means by which the early Nordic success was generalised to
other EU countries in the second generation of mobile telecommunications.

Along these lines, a 1998 feature in Business Week compared the US experience in
second-generation mobile telephony to that of Europe under the heading ”The Cowboys
versus the Committee” (Brull, Gross, and Yang 1998). The article relates that while
Europe was able to settle quickly on one standard, thereby securing a large initial market,
and was able to keep licensing fees low in order to entice manufacturers, the US allowed
four competing digital standards to flourish, none of which could match the huge
subscriber base of GSM. These developments are considered to account for the
subsequent loss of market share by US equipment manufacturers to European rivals
during the second generation of mobile telephony.

The conventional wisdom could be correct. Even so, there is no guarantee that Europe’s
past successes will be replicated in the third generation of mobile telephony. If standards
were indeed the key to European leadership in this field, it must be recognised that many
of the conditions associated with the development of successful first- and second-
generation standards no longer apply.

After the GSM standard was decided upon, the EC introduced far-reaching institutional
changes in the telecommunications sector. Among other things, these reforms have done
away with a ‘closed’ approach to standards-making, led and controlled by monopolistic
PTOs or PTTs that also had regulatory powers within protected national markets. Yet
these conditions may have been largely responsible for the relatively rapid agreement in
Europe on GSM as a single standard for mobile telephony. In the wake of the
institutional changes that accompanied the introduction of GSM – notably, market
liberalisation and ‘open’ standards development -- it may prove far more difficult to
ensure ‘virtuous’ interaction between the supply and demand sides of the mobile
telecommunications market in developing and implementing the third generation
standard, UMTS. As Corrocher (2001) argues, mobile internet, which will be a prominent
part of the third generation, represents a major technological opportunity for Europe. In
this context, though, “the most important and sensitive issue is going to be the design of
appropriate pricing schemes for consumers and business users: in this respect, the role of
public policy is going to be crucial” (ibid.: 57).

In addressing the question of whether it is possible to replicate the success of GSM with a
third generation standard, it may be instructive to consider the track records of other
major second-generation standards. The closed Japanese standard, as noted earlier under
‘geographical boundaries’ (sub-section 7.3), failed to capture any external markets. Yet
the D-AMPS standard, a product of the much more open approach to standards
development in the US-also failed to replicate the success of the first-generation AMPS
standard. D-AMPS sought to exploit the advantage of backwards-compatibility with the
earlier AMPS standard, which had been a global success. However, the de-centralised,
voluntaristic and producer-dominated standard-setting setting regime within which D-
AMPS was developed detracted greatly from its success on international markets. due to
a general lack of agreement and co-ordination within the domestic market. Significantly,
the standard-setting regime that emerged in Europe during the second generation of
mobile telecommunications now resembles its counterpart in the US much more closely
than was the case when GSM was first developed. The prospects for a replication of the
GSM ‘success story’ are therefore not entirely certain. Nevertheless, it should also be
noted that there are some early indications that the European standard, UMTS has
acquired a promising early lead over its rivals, due to collaboration with Japanese
partners (Dalum 2001: 14).

In the second generation of mobile telecommunications, the ‘playing field’ became
significantly larger and so too did the ‘rule-book’. The ‘game’ became considerably
more complicated. It is still not clear whether the EU’s home team will be able to draw
advantage from the new rules and playing conditions.


America's Network. 1996. Wonders of the world phone: L. M. Ericsson's planned
              development of the world phone. America's Network, 46 - 47.
Andersen, S. S., and K. A. Eliassen. 1993. The EC as a new political system. In Making
              policy in Europe -- The Europeifiation of national policy-making, edited by
              S. S. Andersen and K. A. Eliassen. London: Sage publications.
Bach, D. 2000. International co-operation and the logic of networks: Europe and the
              Global System for Mobile Communications (GSM) -- BRIE Working Paper
              no. 19. Berkely: University of California at Berkely.
Becker, F. 1991. PCN: The key to mobility. Siemens Review Telecommunications
              Special:15 - 16.
Bekkers, R., G. Duysters, and B. Verspagen. 2000. Intellectual property rights, strategic
              technology agreements and market structure: The case of GSM. Paper read
              at Swedish International Symposium on Economics, Law and Intellectual
              Property Rights, June, at Stockholm, Sweden.
Bekkers, R., and I. Liotard. 1999. European standards for mobile communications: The
              tense relationship between standards and intellectual property rights.
              European Intellectual Property Rights 3:110 - 126.
Braugenhardt, S., and J. Kaaman. 1991. Open Network Provision -- EGs koncept för
              öppna telenät. Tele (Swedish ed.) 3:36 - 40.
Brull, S. V., N. Gross, and C. Yang. 1998. Cell phones: Europe made the right call.
              Business Week, September 7, 107 -.
Cattaneo, G. 1994. The making of pan-European networks as a path-dependency process:
              The case of GSM versus Integrated Broadband Communication. In Global
              Telecommunication strategies, edited by P. Pogorel. Amsterdam: Elsevier
Clever, M. 1999. Mass market solutions for mobile data. Telecommunications 33 (6):40 -
Commission of the European Communities. 1987. Green paper on the development of the
              common market for telecommunications services and equipment, CEC
              Communication - COM (87) 290 final, 30.6 1987. Brussels-Luxembourg:
———. 1994. Towards the personal communications environment: Green paper on a
              common approach to the field of mobile and personal telecommunications
              in the European Union, CEC Communication - COM (94) 145, 27.4.1994.
              Brussels-Luxembourg: CEC.
Corrocher, N. 2001. The internet services industry: Sectoral dynamics of innovation and
              production and country-specific trends in Italy and the UK: ESSY project
              report on telecommunications. Milan, Italy: CESPRI, Bocconi University.
Dalum, B. 2001. Data communications -- The satellite and tv subsystems: ESSY project
              report, telecommunications paper no. 4. Aalborg, Denmark: Department of
              Business Studies, Aalborg University.

Dalum, B., C. Freeman, R. Simonetti, N. Von Tunzelmann, and B. Verspagen. 1999.
              Europe and the information and communication technologies revolution. In
              The economic challenge for Europe: Adapting to innovation-based growth,
              edited by J. Fagerberg, P. Guerrieri and B. Verspagen. Cheltenham, UK:
              Edward Elgar.
Dalum, B., and G. Villumsen. 2001. Fixed data communications -- Challenges for
              Europe: ESSY project report, telecommunications papers no. 5-6. Aalborg,
              Denmark: Department of Business Studies, Aalborg University.
Dang-Nguyen, G., V. Schneider, and R. Werle. 1993. Corporate actor networks in
              European policy-making: Harmonizing telecommunications policy. Köln:
              Max Planck Institut für Gesellschaftsforschung.
DaSilva, J. S., and B. E. Fernandes. 1995. The European research program for advanced
              mobile systems: Addressing the needs of the European Community. IEEE
              Personal Communications 2 (1):14 - 19.
Dixon, H. 1988. Ericsson switches its global strategy. Financial Times, 25 November, 8.
Dodsworth, T. 1988. Telecoms partners grow apart. Financial Times, 10 June, 8.
Drake, W. 1994. The transformation of international telecommunications standardisation:
              European and global dimensions. In Telecommunications in transition:
              Policies, services and technologies in the European Community, edited by
              C. Steinfeld and Associates.
Emmerson, B. 1998. On the move. Communications International 25 (4):40 - 44.
ETSI. 2001a. Key ETSI standards -- DECT [Internet]. Available:
              2001a [cited January 16 2001a].
———. 2001b. Key ETSI standards -- TETRA [Internet]. Available:
              2001b [cited January 16 2001b].
ETSI Highlights. 1989. News of ETSI's technical committees. ETSI Highlights,
              November, 6.
European Commission. 1996. Introduction to COST co-operation. Luxembourg: Office
              for Official Publications of the European Communities.
European Commission. 2001. Job opportunities in the information society: Exploiting the
              potential of the information revolution -- COM (1998) 590 final - EN.
              1998 [cited 03/01 2001].
FINTECH. 1990. Untitled. Mobile Communications (March).
———. 1991. Untitled. Mobile Communications (March).
Fridlund, M. 1998. Switching relations: The government development procurement of a
              Swedish computerized electronic telephone switching technology. In ISE:
              Innovation systems and European integration [CD-ROM], edited by C.
              Edquist and F. Texier. Linköping, Sweden: Systems of Innovation Research
              Programme, Department of Technology and Social Change, Linköping
Funk, J. L. 1998. Competition between regional standards and the success and failure of
              firms in the world-wide mobile communication market.
              Telecommunications Policy 22 (4/5):419 - 441.
Funkschau. 1993. Motorola/Nokia: Austach von GSM-patenten. Funkschau, 50.

Garrard, G. A. 1998. Cellular communications: Worldwide market development. Boston:
               Artech House.
Glimstedt, H. 2000. The competitive dynamics of technological standards: The case of
               the third generation in cllular communications. Paper read at Seminar for
               the Systems of Innovation Research Programme, Department of
               Technology and Social Change, Linköping University, December, at
               Linköping, Sweden.
Granstrand, O., E. Bohlin, C. Oskarsson, and N. Sjorberg. 1992. External technology
               acquisition in large, multi-technology corporations. R&D Management
               22:111 - 133.
Granstrand, O., P. Patel, and K. Pavitt. 1997. Multi-technology corporations: Why they
               have 'distributed' rather than 'distinctive core' competencies. California
               Management Review 39 (4):8 -25.
Haddon, A. 1991. DCS1800: The standard for personal communication networks.
               Telecommunications (June):61 - 63.
Hardiman, E. 1990. The role of private mobile radio. Paper read at World Mobile
               Telecommunications Conference, September 24 - 25, at London.
Hawkins, R. 1992. The doctrine of regionalism: A new dimension for international
               standardisation in telecommunications. Telecommunications Policy
               (May/June):339 - 353.
Hawkins, R., R. Mansell, and W. E. Steinmueller. 2000. Liberalization and the process
               and implications of standardization. In Mobilizing the information society:
               Strategies for growth and opportunity, edited by R. Mansell and W. E.
               Steinmueller. Oxford/New York: Oxford University Press.
Hawkins, R. W. 1995. Enhancing the user role in the development of technical standards
               for telecommunications. Technology Analysis and Strategic Management 7
               (1):21 - 40.
Hollingsworth, J. R., P. C. Schmitter, and W. Streeck. 1994. Capitalism, sectors,
               institutions, and performance. In Governing capitalist economies, edited by
               J. R. Hollingsworth, P. C. Schmitter and W. Streeck. New York: Oxford
               University Press.
Holst, G.-M., ed. 1994. The Teldok yearbook - 1994, Teldok report 86E. Stockholm: Hj.
               Brolins Offset AB.
———, ed. 1997. The Teldok yearbook - 1997, Teldok report 116. Stockholm: Hj.
               Brolins Offset AB.
———, ed. 1999. Teldoks Årsbok - 2000, Teldok report 130. Stockholm: Hj. Brolins
               Offset AB.
Iversen, E. J. 2000. Standardisation and intellectual property rights: Conflicts between
               innovation and diffusion in new telecommunications systems. In
               Information technology standards and standardisation: A global
               perspective, edited by K. Jacobs. Hershey, USA/London, UK: IDEA Group
Johnston, W. 1998. Europe's future mobile telephone system. IEEE Spectrum 35 (10):49
               - 53.
Karlsson, M. 1998. The liberalisation of telecommunications in Sweden -- Technology
               and regime change from the 1960s to 1993. Ph.D. Dissertation, Department

               of Technology and Social Change, Linköping University, Linköping,
Law, C. E. 1991. Personal communications networks: What do they mean? Mobile and
               Satellite Single Market Review 1.
Lernevall, S., and B. Åkesson. 1997. Från myndighet till bolag 1966 - 1993, Svenska
               Televerket Del VII. Stockholm: Telia.
Lilliesköld, G. 1989. NMT, GSM och 'sladdlös' PABX. Modern Elektronik (3):35.
Lindmark, S. 1995. The history of the future: An investigation of the evolution of mobile
               telephony. Licentiate Thesis, Department of Industrial Management and
               Economics, Chalmers University, Gothenburg, Sweden.
Lindmark, S., and O. Granstrand. 1995. Technology and systems competition in mobile
               telecommunications. In Beyond competition: The future of
               telecommunications, edited by D. M. Lamberton. Amsterdam, The
               Netherlands: Elsevier Science B. V.
McKelvey, M., F. Texier, and H. Alm. 1998. The dynamics of high-tech industry:
               Swedish firms developing mobile telecommunications systems. In ISE:
               Innovation systems and European integration [CD-ROM], edited by C.
               Edquist and F. Texier. Linköping, Sweden: Systems of Innovation Research
               Programme, Department of Technology and Social Change, Linköping
Meurling, J., and R. Jeans. 1994. Mobiltelefon - En idé som skapade en världindustri.
               Stockholm: Informationsförlaget.
———. 1995. A switch in time: AXE - Creating a foundation for the information age.
               London: Communications Week International on behalf of Ericsson
Microcell. 1991. ETSI eyes new PMR specs, runt into DECT LAN flap. Microcell,
               October, 6.
Muller, J. 1992. Competition and regulation in telecommunications. Telecommunications
               Policy 19 (2):161 - 164.
Mölleryd, B. 1997. Så byggdes en värdsindustri -- Entrepreörskaps betydelse för svensk
               mobiltelefoni. Stockholm: Teledok 28.
Nissinen, M., and P. Niskanen. 1999. COST -- Scientific co-operation on researchers'
               terms: A study of Finnish participation. Espoo, Finland: VTT (Technical
               Research Centre of Finland).
NUTEK. 1997. Telecommunication: Evaluation of the research programme, 1993 - 1996.
               Stockholm: NUTEK.
OECD. 1999a. Communications outlook -- 1999. Paris: OECD.
Paetsch, M. 1993. Mobile communications in the US and Europe: Regulation, technology
               and markets. Boston/London: Artech House.
Petersson, O., and D. Söderlind. 1992. Förvaltningspolitik. Stockholm: Publica.
Pisjak, P. 1994. Interdependence between regulation and technological innovation in the
               telecommunications sector. Technology Analysis and Strategic
               Management 6 (3):289 - 303.
Pohajakallio, P. 1998. Casting the cellular net. Telecommunications 32 (4):47 - 50.
             Post och Telestyrelsen. 1998. Konkurrenssituationen på olika delmarknader
             inom telekommunikationsområdet, 1997 [Report on-line, cited 17 August,

             1999]. Stockholm: National Post and Telecom Agency [Available from
Prins, C., and M. Seissel. 1993. The new European Telecommunications standards
               Institute policy: Conflicts between standardisation and intellectual property
               rights. European Intellectual Property Rights.
             PwC, and Öhrlings-Price-Waterhouse-Coopers. 1999. Den svenska
             marknaden för telekommunikation, 1998 [Report on-line, cited 17 August,
             1999]. Stockholm: National Post and Telecom Agency [Available from
Ramsdale, P. 1991. PCN implementation in the UK. IEE Review (February):76 - 77.
———. 1991. PCN infrastructure: Looking at the options. Communications International
               (November):53 - 56.
Ruotto, A. 1998. Governance within the European television and mobile communications
               industries: PALplus and GSM - A case-study of Nokia. Ph.D. Dissertation,
               University of Sussex, Sussex, UK.
Sandholtz, W. 1993. Institutions and collective action: The new telecommunications in
               Europe. World Politics 45 (2).
Schreiber, K. 1991. The new approach to technical harmonisation and standards. In The
               state of the European Community: Policies, institutions and decisions in the
               transition years, edited by L. Hurwitz and L. Lequesne.
Shankar, B. 1997. Airing the standards debate. Telecommunications 31 (10):41 - 44.
Taylor, R., and T. Dixon. 1989. Ericsson and GE in mobile phone venture. Financial
               Times, 25 August, 24.
Temple, S. R. 1992. The European Telecommunications Standards Institute -- Four years
               on. Electronics and Communication Engineering Journal 4 (4):177 - 181.
TETRA MoU. 2001. TETRA -- Facts [Internet]. Available:
      2001 [cited January 16 2001].
Tuckett, R. 1993. ETSI's IPR policy: The implications for companies using European
               telecoms standards. Patent World 23 (September).
Wallenstein, G. 1990. Setting global telecommunications standards.
Wells, P. E., and P. Cooke. 1991. The geography of international strategic alliances in the
               telecommunications industry: The cases of Cable and Wireless, Ericsson,
               and Fujitsu. Environment and Planning A 23:87 - 108.
Wilkinson, S. L. 1991. They're stealing our diamonds: The standards assault on patents.
               Canadian Intellectual Property Review 8 (2):193 - 197.


Shared By:
gjmpzlaezgx gjmpzlaezgx