Telecommunications Policy 27 (2003) 325–332
Standards in wireless telephone networks
*, David Salant
, Leonard Waverman
Department of Public Policy, Tel Aviv University, 69978 Tel Aviv, Israel
CEPR, London, UK
NERA Economic Consulting, San Francisco, USA
London Business School, London, UK
Since 1994, Europe and North America have taken divergent approaches to managing spectrum for
wireless voice and data services, the so-called 2G and 3G bands. (There are several so-called generations in
mobile—ﬁrst generation (1G) is analogue service, second generation (2G) is digital, while third generation
(3G) refers to higher bandwidth packet switched networks; 2.5G refers to upgrading a 2G network to
permit near 3G data rates.) The European Community has mandated a harmonized standard, GSM, in the
2G bands, and has adopted Wideband CDMA (WCDMA) in the 3G bands. In contrast, the North
American approach has been to allow the market to decide, that is, operators have been free to choose
among the recognized four digital wireless standards for 2G: CDMA/IS-95, GSM, TDMA and iDEN.
The issue of market-based versus mandated standards has been addressed in many other industries. In
most settings where network effects are present, compatibility across platforms (i.e., standardization) has
been a key determinant of the success or failure of a particular technology. In the case of wireless
telecommunications, however, interconnection and the availability of the relevant infrastructure can be a
substitute for compatibility. An individual subscribing to any one of the wireless technologies in the US can
easily make calls to and receive calls from subscribers to any one of the other standards (or to and from the
wire-line POTS network) as long as there is (i) interconnection between networks and (ii) the relevant
infrastructure is in place. In the US (and several other developed countries), interconnection has been
achieved by standard interconnection protocols.
In this paper, we discuss the tradeoff between mandated standards and market-driven standards in the
wireless telecommunications industry. We discuss the theoretical advantages of each approach, and provide
institutional background on the developments of 1G–2G and discuss the implications of our results for the
current debate about 3G standards.
r2003 Elsevier Science Ltd. All rights reserved.
*Corresponding author. Eitan Berglas School of Economics, Tel Aviv University, 69978 Tel Aviv, Israel. Tel.: +972-
3-640-6742; fax: +972-3-640-7382.
0308-5961/03/$ - see front matter r2003 Elsevier Science Ltd. All rights reserved.
Europe and North America have taken divergent approaches to managing spectrum especially
for wireless voice and data services. The US, Canada and other developed countries have allowed
the market to determine standards. In contrast, the European Community (EC) has relied on
mandated standards set by the European Telecommunications Standards Institute (ETSI). These
decisions have likely had an impact on prices, penetration rates, service and feature availability,
new technology development and deployment and coverage of 2G services in the EC and the US.
The decisions likely will also have an impact on the deployment, price and availability of the next
generation of wireless voice and data services, the so-called third-generation technologies.
Many have argued that the EC adoption of a uniform 2G/GSM standard is one of the great
successes of European telecommunications policy, and the North American regulators’ decision
to let the market determine standards is a great failure.
This paper seeks to provide the ﬁrst step towards examining the economics underlying these
views. In particular, we discuss how beneﬁts from wireless telecommunications are affected by (i)
market-determined standards versus mandated standards and (ii) compatibility and coverage. By
compatibility we mean the ability of those operating different networks to use the same
equipment. For wireless technologies, this means that handsets or terminals, as well as base
station and other network equipment can work on all compatible networks. Coverage refers to the
area over which the consumer can use his or her handset.
Section 2 brieﬂy describes the history of standard setting for wireless telecommunications in
Europe and the US. Section 3 discusses the economics of standard setting as applied to wireless
telecommunications. Section 4 provides further discussion.
2. History of spectrum management regulation
2.1. North American market
In the early 1980s the Federal Communications Commission (FCC) approved the Analog
Mobile Phone System (AMPS) as a uniform standard. The FCC allocated 2 25 MHz of
spectrum in each metropolitan statistical or service area (MSA) and each rural service area (RSA).
This allocation process took several years. The incumbent local exchange carriers (ILECs)
received licenses in their home markets. The other license in each MSA/RSA was assigned ﬁrst by
comparative hearings, and then via lotteries. This was the only allocation of spectrum for mobile
phone services in the USA until 1994.
In 1994, the FCC allocated 120 MHz (or 2 60 MHz) of new spectrum for digital or Personal
Communications Services (PCS) licenses. In September of that year, the FCC decided to divide
this spectrum into six frequency bands in each region. In December 1994, the FCC began the ﬁrst
of a series of auctions for broadband PCS licenses.
The ﬁrst auction, for two 2 15 MHz licenses,
The FCC had previously auctioned off frequency for narrowband PCS licenses. Narrowband is smaller blocks, from
50 kHz to 2 50 kHz, as compared to broadband PCS, which has been allocated in blocks of sizes ranging from
25 MHz to 2 15 MHz. Narrrowband is used for advance paging and messaging services. Broadband is primarily
used for voice and other data services.
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332326
the A and B blocks, in each of 51 major trading areas (MTAs), began in December 1994 and
ended in March of 1995.
Subsequently, the FCC auctioned one 2 15 MHz C block license in each of 493 basic trading
areas (BTAs) in 1996, and three 2 5 MHz D, E and F block licenses in each of the 493 BTAs in
an auction completed in January of 1997.
The FCC has also allowed the AMPS license holders to re-farm their spectrum, i.e., convert it
from analog to digital service. This re-farming began in 1995, and much of the AMPS spectrum
has, by now, been converted.
In comparison to its policy of the early 1980s in the AMPS cellular bands, the FCC did not
mandate a speciﬁc standard for the PCS bands. Carriers were free to choose whatever standard
they wished. Today, there is nearly equivalent nationwide coverage in the US for CDMA, TDMA
and GSM. In most cases, coverage has been achieved with a combination of re-farmed AMPS
frequencies and the newer PCS frequencies. This coverage has been achieved without the FCC (or
any other regulatory body) mandating a standard and without a mandate for nationwide
roaming. Market forces led to this outcome.
2.2. The European market
The deployment of digital service ﬁrst occurred in Europe. Western Europe saw rapid growth in
analog mobile telephony service during the 1980s. Most countries had a single system that was not
always compatible with the systems of its neighbors. Hence, subscribers had difﬁculties roaming.
The EU countries elected to try to introduce a harmonized standard for 2G or digital service.
The belief was that apart from ensuring roaming, a uniform standard would permit greater
economies of scale in equipment supply.
In 1982, the Conf!
erence des Administrations Europ!
eenes des Postes et T!
(CEPT) decided that a new digital standard be developed to cope with the increasing demands on
European mobile networks. CEPT established a working party to develop a set of common
standards for a pan-European cellular network. This working party was known as the Group
Speciale Mobile (GSM). CEPT identiﬁed the importance of the availability of common spectrum
in the development of a European system and made representations to the European Commission
on this issue. This resulted in the EC issuing a directive under which European States were
required to set aside spectrum in the 900 MHz band for the future development of a European
mobile telecommunications system.
The technology adopted was the Global System for Mobile Communications (and this took
over the acronym GSM). In 1987, the operators from the CEPT countries signed a Memorandum
of Understanding, usually referred to as the GSM MoU, in which they agreed to deploy the GSM
standard at the same frequency in order to facilitate roaming.
In 1989, CEPT transferred the
GSM committee to the ETSI, itself formed in 1988. ETSI completed the speciﬁcations of the
Not all of the C block and F block spectrum have been used due to post-auction defaults, subsequent re-auctions
and ongoing litigation.
Network operators from 13 countries signed the ﬁrst Memorandum of Understanding on 7 September 1987. They
committed to licensing GSM technology by 1 January 1991.
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332 327
By 1989, Germany had already awarded a GSM concession, the D2 license. By the end of 1993
there were already more than 1 million GSM users in Europe. By contrast, in the US, the FCC did
not even complete its ﬁrst auction allocating PCS spectrum until March of 1995. By that time,
every country in Europe had at least one GSM operator.
2.3. Third-generation wireless technologies (3G)
Over the past few years, there has been ongoing controversy about the likely migration to 3G.
Two main standards have been proposed, UMTS also known as Wideband CDMA (WCDMA)
and CDMA2000. Both use coding schemes derived from the second-generation version of
CDMA, also called CDMA-One. There are a number of differences between CDMA2000 and
WCDMA. However, CDMA2000 is a natural migration from CDMA-One, while WCDMA is
essentially incompatible with any existing technology.
Most interested European interested parties, i.e., operators, equipment suppliers and policy
makers, seem to favor WCDMA. The reason for their favoring WCDMA is possibly due to the
fact that it puts them on even footing with CDMA 2G equipment suppliers and operators whereas
CDMA2000 would give the latter a competitive cost and time to market advantage. In May 1998,
European Union telecommunications ministers endorsed the draft law designed to keep the 15-
nation bloc as one with regard to 3G. The EC mandated deployment of 3G services by the end of
2002. No operator in the EC member countries met this target deployment date for WCDMA.
In North America, two nationwide networks and a number of regional carriers have deployed
CDMA2000 on their networks. Most new terminals and most new additions to these networks are
CDMA2000. In Asia, SK (South Korea) Telecom launched CDMA2000 in October 2000. By
May 2001, LG Telecom and SKF had followed suit.
Perhaps the most interesting developments have occurred in Japan. NTT DOCOMO launched
WCDMA in late 2001, and J-Phone did the same almost a year later. KDDI launched
CDMA2000 in April 2002. By the end of January 2003, KDDI had signed up over 5.3 million
CDMA2000 subscribers and subscriptions to NTT DOCOMO (for WCDMA) had fallen.
DOCOMO recently stated that it would be unlikely to reach its goal of 320,000 subscribers to
WCDMA by 2003. Despite the push of WCDMA from Europe, by the end of January 2003,
CDMA2000 had signed up almost 32,000,000 subscribers worldwide, compared with 160,000 for
CMDA2000 is a relatively simple upgrade of existing CDMA technology, and so most, if not
almost all, CDMA operators are deploying it. An increasing percentage of new terminal and
network equipment is now CDMA2000. In contrast, WCDMA is essentially incompatible with
GSM (TDMA). It has enough differences with CDMA2000 that the development of network and
terminal equipment is likely taking more time than most anticipated. These limitations no doubt
are part of the explanation of why WCDMA’s introduction has been delayed in most places, and
in the few places where a WCDMA system has been introduced, why it has failed to attract
signiﬁcant numbers of new subscribers.
If CDMA2000 succeeds and WCDMA does not, or if CDMA2000 succeeds much more quickly
than WCDMA, any EU policy that may have assumed WCMDA or locked operators into
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332328
WCMDA may prove to be a very costly public policy decision. So, not only is there a historical
question about the relative advantages of the European and North American approaches to
standardization policy for 2G, but also there is a current policy question about the 3G bands as well.
3. The economics of standardization policy as applied to wireless standards
Broadly speaking, there are three ways that standards get set in practice: (I) De facto standards,
i.e., standards set primarily by the market. These standards are often proprietary. (II) Voluntary
industry agreements, where standards are often jointly developed. These standards are typically
open standards, that is, they are not proprietary. (III) Standards imposed by National Standards
Bodies (NSBs), or agreed upon by regional or international standards development organizations
There is a relatively large theoretical literature on the economics of standards and
compatibility. The literature is primarily concerned with the private and social incentives to
obtain compatibility, i.e., standardization.
To a large extent, this literature assumes that compatibility (i.e. standardization) has beneﬁts
that make it a socially desirable outcome. This is less obvious in the case of wireless standards.
Here interconnection and coverage are likely more important than a single standard. In the case
of wireless telecommunications, as long as the networks are interconnected and coverage is good,
presumably compatibility should not matter. In addition to quality and price, consumers
presumably care about the size of the network for which they can place and receive calls.
Interconnection ensures that a subscriber on one ﬁxed or mobile network can communicate with a
subscriber on any other ﬁxed or mobile network. Compatibility will matter when the coverage of
the competing network operators differs signiﬁcantly.
When the EC mandates a standard, and North America does not, the EC standard will achieve
coverage within Europe that might not be effectively matched in the US or North America by
DAMPS/TDMA, CDMA, or GSM. However, this advantage can work to the disadvantage of
consumers in the longer run, if the mandated technology is inferior to one used elsewhere.
A key argument in favor of compatibility is that the larger economies of scale in the production
of both terminals/handsets and network infrastructure equipment reduce costs and increase
availability. That is, the higher the sales of chipsets, terminal, and network equipment, the lower
the unit cost. Additionally, the variety of terminal equipment (handsets) tends to be greater.
There are beneﬁts from multiple (competing) standards as well since the types of services tend
to differ across technologies. For example, CDMA networks have offered more and better data
See Gandal (2002).
David and Greenstein (1990) provide a comprehensive survey of earlier work, while Farrell and Klemperer (2003)
provide a detailed survey of more recent work. Gilbert (1992),Katz and Shapiro (1994),Gandal (1995),Matutes and
Regibeau (1996), and Gandal (2002) provide selective reviews of the literature.
A small but growing literature has empirically (statistically) found evidence that consumers value compatibility. See
Greenstein (1993),Gandal (1994),Brynjolfsson and Kemerer (1996), and Gandal, Greenstein, and Salant (1999) for
empirical evidence of network effects in the computer software industry. Other papers that provide empirical evidence
that consumers value compatibility include Saloner and Shepard (1995), the ATM industry, and Gandal, Kende, Rob
(2000), the CD industry.
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332 329
services than were available on GSM networks. Other advantages to market competition include
more technological competition and greater price competition (at least early on) among
competing incompatible standards.
Equipment vendors often have an interest in one standard over another. Such was the case with
CDMA for 2G networks where QUALCOMM and, to a lesser extent, some of the main North
American equipment suppliers favored CDMA over GSM, while European equipment suppliers,
including Nokia, Siemens and Ericsson, favored GSM over CDMA. These preferences hold for
3G as well. Given the vested interests, it is likely that the main telecommunications equipment
vendors would want to tip the standard decision one way or the other.
The economic theory of tipping would suggest that the early adoption of one standard or the
decision to formally set one standard in the EC can tip the whole world toward that standard.
Then the adoption of a single standard by a few large ﬁrms will likely tip the entire market toward
In market competition between wireless standards, interconnection may mean that
the standard tipping results may apply only if one standard gets far out in front of a competing
standard early on before the competing standard has a chance to get established. In the case of
second-generation wireless systems, CDMA succeeded despite the initial lead of GSM.
3.1. The ETSI standardization process in detail
As noted above, the EU mandated adoption of the ETSI standard as mandatory. We now
brieﬂy examine the standardization process within ETSI. ETSI standards are based on the
principle of consensus but with weighted voting (based on European turnover) and a 71% rule for
agreement. Weighted voting based on European turnover means that the process favors European
manufacturers. A small minority can apparently, with relatively few allies, push through virtually
any standard. Thus, ETSI in 1999 approved the UMTS/WCDMA standard well before any
technology was available.
Some US ﬁrms, most notably QUALCOMM, which has no signiﬁcant
revenue in Europe, were not in favor of a mandated harmonized UMTS standard. QUALCOMM
appealed to both US and EU regulators. As a result the EC recanted slightly and belatedly to
permit any IMT-2000 standard and not just WCDMA/UMTS.
So far, there have been virtually no practical implications from this recanting, with the possible
exception of helping lead to technology licensing agreements. The reason is that most European
telecom operators supported WCDMA as an evolution from GSM.
4. Further discussion
In the EC, all digital second-generation networks (2G) deploy the mandated Global System for
Mobile Communications (GSM) standard. In contrast, in North America, and several other
developed countries, standards were determined via the market and there are multiple digital
See Arthur (1989),Cabral and Riordan (1994), and Salant (1991).
The US standards process, on the other hand, is totally open. There is no obligation for ﬁrms to have US earned
revenue, so that European and Asian ﬁrms participate equally.
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332330
A question of interest is whether mandated standards have been more beneﬁcial than standards
determined by the market in the case of wireless telecommunications. The ETSI process for 3G is
somewhat reminiscent of the HDTV standard setting process in which both Japan and the
Europeans locked into an HDTV standard at a very early stage. As Farrell and Shapiro (1992)
note (p. 25) this ‘‘highly centralized approach foregoes the beneﬁts of competition in research and
developmenty’’. The US approach in contrast to the case of HDTV was to encourage
competition for a period of time before deciding on a standard. It is generally agreed that the US
approach of mixing market competition with elements of centralized decision making led to a
better outcome for the US in terms of the HDTV prototype.
Whether mandated standards have been more beneﬁcial than standards determined by the
market in the case of wireless telecommunications depends upon several factors including whether
market competition led to technological improvements in wireless technology, whether
compatibility (standardization) matters for the adoption of wireless technologies, as well as
other regulatory decisions about factors such as calling party pays, roaming and call termination.
We leave these issues for future research.
Arthur, W. B. (1989). Competing technologies, increasing returns, and lock-in by historical events. Economic Journal.
Brynjolfsson, E., & Kemerer, C. (1996). Network externalities in microcomputer software: An econometric analysis of
the spreadsheet market. Management Science,42, 1627–1647.
Cabral, L., & Riordan, M. (1994). The learning curve, market dominance and predatory pricing. Econometrica,62(5),
David, P., & Greenstein, S. (1990). The economics of compatibility standards: An introduction to recent research.
Economics of Innovation and New Technology,1, 3–41.
Farrell, J., & Klemperer, P. (2003). Handbook of industrial organization chapter on lock-in through switching costs
and network effects, forthcoming.
Farrell, J., & Shapiro, C. (1992). Standard setting in high-deﬁnition television. Brookings Papers: Microeconomics (pp.
Gandal, N. (1994). Hedonic price indexes for spreadsheets and an empirical test for network externalities. RAND
Journal of Economics,25, 160–170.
Gandal, N. (1995). A selective survey of the literature on indirect network externalities. Research in Law and Economics,
Gandal, N. (2002). Compatibility, standardization, & network effects: Some policy implications. Oxford Review of
Economic Policy,18, 80–91.
Gandal, N., Greenstein, S., & Salant, D. (1999). Adoptions and orphans in the early microcomputer market. Journal of
Industrial Economics,XLVII, 97–106.
Gandal, N., Kende, M., & Rob, R. (2000). The dynamics of technological adoption in hardware/software systems: The
case of compact disc players. RAND Journal of Economics,31, 43–61.
Gilbert, R. (1992). Symposium on compatibility: Incentives and market structure. Journal of Industrial Economics,40,
Greenstein, S. (1993). Did installed base give an incumbent any (measurable) advantages in federal computer
procurement. RAND Journal of Economics,24, 19–39.
Katz, M., & Shapiro, C. (1994). Systems competition and network effects. Journal of Economic Perspectives,8, 93–115.
See Farrell and Shapiro (1992).
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332 331
Matutes, C., & Regibeau, P. (1996). A selective review of the economics of standardization: Entry deterrence,
technological progress, and international competition. European Journal of Political Economy,12, 183–206.
Salant, D. (1991). Some Stochastic Oligopoly Races for Experience. GTE Laboratories.
Saloner, G., & Shepard, A. (1995). Adoption of technologies with network externalities: An empirical examination of
the adoption of automated teller machines. RAND Journal of Economics,26, 479–501.
N. Gandal et al. / Telecommunications Policy 27 (2003) 325– 332332