CHAPTER 3

The telecommunications industry in Europe up to 2000

‘The great advantage [the telephone] possesses over every other form of electrical apparatus consists in the fact that it requires no skill to operate the instrument.’

– Alexander Graham Bell, 1878

Less than a year after Bell was granted his first patent, the telephone arrived in Europe. In Sweden the first phone line had been laid in 1877 by Kongliga Elektriska Telegraf Verket (Royal Electric Telegraph Works). In 1883, three years after the first exchange was installed in Stockholm, Swedish engineer Henrik Tore Cedergren formed a company called Stockholms Allmanna Telefonaktiebolag (Stockholm Public Telephone Company). Cedergren’s ambition was to put a phone in every household in Sweden, and he began collaborating with the new company Lars Magnus Ericsson had formed.

In the German empire, in 1877 all telephone operations in the country were declared to be under the control of the postal authority, while in Russia the government issued a decree in 1881 that the development of an urban telephone network should be permitted, with the state becoming increasingly involved from 1884 onwards.

In 1882 Britain’s postmaster general, Henry Fawcett, gave telephone companies the right to establish public telephone stations in both private and public places. New ‘call offices’ were soon to be found in ‘silence cabinets’ in shops, railway stations and many other public locations.

In 1883 a Scottish engineer at the National Telephone Company’s Glasgow office patented an automatic telephone exchange, which allowed a subscriber at a branch exchange to be connected to the central exchange without manual intervention at the branch level. This was a big step towards one of the biggest single steps in the history of the industry – the invention of a fully automatic exchange.

In 1896 Italian electrical engineer Guglielmo Marconi visited the British Post Office, wanting to talk about a new system of ‘telegraphy without wires’ that he’d invented. Marconi had already approached the relevant bodies in his native Italy but they hadn’t been interested.

The British engineer-in-chief was, however, and by 1899 Marconi’s wireless systems were powerful enough to cross the English Channel, a distance of some 32 kilometres (20 miles). Two years later the engineering had advanced sufficiently to allow a signal to be transmitted across the North Atlantic, from Cornwall to Newfoundland.

By the beginning of the 1900s the concept of the telephone had firmly established itself, at least in the developed world. The industry looked set to continue to grow, with many businesses using it in preference to the older technology of telegraphy.

Telecoms technology continued to improve at an accelerating pace. In 1926 Paul M Rainey of Western Electric made and patented a facsimile machine that converted images into digital signals, using a technique called pulse code modulation (PCM). It’s the key digital technology that enabled compact disks to be made and also features in computer disks, digital audio and telephony, and many other modern applications. Ten years later British engineer Alec Reeves invented a means of using PCM to sample and encode telephone signals.

In 1931 the world’s first microwave link was demonstrated over the 32 kilometres (20 miles) of the English Channel that separates Dover and Calais. Microwave technology benefited hugely from developments during the Second World War (1939–45), especially in the field of radar.

By the end of the 1930s telephone instruments were generally reliable, transmission systems could reach across oceans, both wirelessly and through cable, and automatic switching was being widely adopted in the developed world. Operating efficiency was enhanced by the use of ‘multiplexing’, a technique by which the capacity of a transmission link could be optimised, and transmission technology also benefited from research into different areas of the electromagnetic spectrum.

Over the next few years, further digital techniques were developed that would lead to a complete change in the way telephone exchanges worked. During the war years, Alan Turing and Tommy Flowers developed the world’s first digital computer, designed to decipher codes generated by German Enigma encryption machine.

In 1956, Britain’s Post Office began work on a fully digital exchange. It produced a prototype 800-line exchange, which was brought into service at the Highgate Wood exchange. While the new technology worked reasonably satisfactorily, the experimental exchange itself was deemed a failure, and the Post Office backed away from the all-digital development and chose to adopt older technologies, such as crossbar. As a result, from having been the world’s largest exporter of telecommunications equipment before the war and the technological leader after, Britain fell down the rankings year after year throughout the 1950s, ’60s and ’70s.

In 1965 Charles Kao and George Hockham at STC’s UK research laboratory realised that the high attenuation in fibres (a reduction in the strength of a signal over distance, which limited the range and the quality of the signal, rendering it essentially impractical) was not the result of some immutable physical law, but the consequence of impurities in the material. The American company Corning Glass subsequently produced ever-higher-quality material with ever-lower attenuation, to the point where a fibre-optic cable wasn’t only comparable to the best coaxial copper cables, but far superior. By 1977, when the first commercial fibre cable was laid (in Turin, Italy), fibre cables were capable of transmitting far greater bandwidths than copper, over longer distances, before (the inevitable) attenuation meant that reamplification was required. Fibres of that vintage could manage anything up to 150 kilometres (90 miles), compared to copper’s limit of 30 kilometres (20 miles), and they were also cheaper to manufacture.

Over the 40 or more years since, the capability of fibre systems has improved almost beyond belief. Much of the best work has been undertaken by Nippon Telegraph and Telephone (NTT) Corporation, Japan’s largest telecommunications company. In 2006 it set a new record when it managed to transmit a signal at a rate of 14 Terabits per second (Tbps) (14,000,000,000,000bps) over a 160-kilometre (100-mile) fibre. Bell Labs bettered this in 2009, with a 15.5Tbits transmission over 7,000 kilometres (4,350 miles). A year later NTT was back in the record books with 69.1Tbits over 240 kilometres (150 miles), and two years later it had raised the rate to 1 Petabits per second (Pbps) over a 50-kilometre (30-mile) fibre. By 2017 the distance such a signal could travel unamplified had quadrupled to 200 kilometres (125 miles). (These references to Tbits and Pbits are somewhat abstract for most people, but in simple terms a fibre with a 1Pbps capacity could transmit 10,000 hours of high-definition television video in a single second.)

On 25 September 1956 the first transatlantic submarine cable opened for business, linking Gallanach Bay in Scotland with Clarenville in Newfoundland. The cable, named TAT-1 (Transatlantic No. 1), was a collaborative effort between AT&T (50%), the British Post Office (40%) and the Canadian Overseas Telecommunications Corporation (10%). It began life with a capacity of 36 channels (enabling 36 simultaneous voice calls to be made). Together, on its first day alone, these carried nearly 600 calls between the UK and the US and more than 100 from the UK to Canada.

Analogue cables such as this remained in production well into the 1980s, before eventually being replaced by fibre-based systems. Demand for international calls – especially across the Atlantic – continued to rise throughout the 1950s, ’60s and ’70s, and over the years TAT-1 was joined by TAT-2, TAT-3 and so on, until TAT-7, the last analogue cable, was laid in 1983. This had capacity for 4,000 simultaneous calls, or more than 100 times the capacity of the first in the series.

Laid in 1988, TAT-8 was the first transatlantic submarine cable to use fibre-optic technology. It had 10 times the capacity of its immediate predecessor (40,000 telephone channels or 280Mbps of data) but wasn’t as reliable as had been hoped – shark attacks on the cable invariably ended badly for all concerned, with the sharks being electrocuted and the cable stopping working. Subsequent cables were laid with interference shielding, after which shark and system coexisted happily.

Privatisation in Europe

At the end of the 1980s, almost all of the telecommunications operators in Europe were still under state ownership. Britain had privatised Cable & Wireless and BT (in 1981 and 1984, respectively) but it would be another eight years before the next European privatisation.

In 1992 the Italian state sold about a quarter of the stake it held in Società Torinese per l’Esercizio Telefonico, the holding company that owned the five separate companies that would later be brought together as Telecom Italia.

No fewer than five PTTs were privatised in 1995. First, the Netherlands sold a 30% interest in Koninklijke KPN, further reducing its holding to 45% later that year and then to 35% in 2000.

Spain’s Sociedad Estatal de Participaciones Patrimoniales reduced its shareholding in Telefónica by 12%, leaving it with just 19.85%. A second disposal reduced this to 1.9% and by the end of that year the company was entirely privately owned.

In 1997 the government of Denmark sold 49% of its stake in Tele Danmark A/S (TDC). Later that year Ameritech acquired a 34.4% stake in TDC, and the government disposed of its residual stake, selling a further 7.2% to Ameritech and the balance to private investors.

The Portuguese government offered 26.3% of Portugal Telekom to private investors in 1995. A second sale a year later took the state’s interest down to 51%, and by 1997 this holding was nearly halved, leaving just 25.1% in government hands.

Finally, in 1995 the Belgian government sold 49.9% of Belgacom to the ADSB consortium consisting of Ameritech, TDC and Singapore Telecom, effectively transferring about 20% of the company from the Belgian state to the Danish and Singaporean states. ADSB eventually sold its interest in 2004, when Belgacom shares were listed on the Brussels Stock Exchange.

In 1996 the government of Greece sold 8% of the Hellenic Telecommunications Organisation to private investors. Further disposals in 1997 reduced the government holding to 80% and a year later to 65%. Sales in 1999 and 2001 reduced this to 40%.

Telekom Austria was separated from the Post Office in 1996 and privatised in the same year when the state sold 25% of the equity to Telecom Italia, which had already bought 25% of Telekom Austria’s Mobilkom mobile subsidiary. Telecom Italia increased its stake to 29.8% when the company was listed on the Vienna Stock Exchange in 2000, but offloaded 15% of this in late 2002, before finally selling out in early 2004.

France Telecom was privatised in 1997, when 25% of the company’s shares were offered to private investors. A second tranche was sold in 1999, taking the state’s interest down to 63.6%. Further disposals were made in 2004, leaving the government with less than 50%, and by the end of 2005 more than two thirds of the company’s shares were in private hands.

Swisscom was privatised in 1998, when 34.5% of the company was sold to investors.

Ahead of privatisation in 1988, Telecom Finland changed its name to Sonera Corporation. The Finnish government sold an initial 22.3% interest in the business at this point, and disposed of a further 20% a year later. Following Sonera’s merger with Sweden’s Telia in 2002, the Finnish state’s stake in the enlarged business dropped to 19%. Telia itself remained in the hands of the state until 2000, when 29.4% of the company was sold to investors. Following the Sonera merger, that dropped to 45.3%.

Telenor, the Norwegian PTT, was one of the last to be privatised. In 2000 the government offered a 22% stake in the business to private investors, this being followed by a second smaller sale in 2003, which took the government’s stake down to 62.5%. A final tranche was made available the following year, further reducing the government’s interest to 54%, which is where it stands today. Telenor remains the only European PTT where the state retains a majority stake in the business.

The first European mobile network was launched in Sweden by Comvik in 1981, one month ahead of a similar launch by Televerket (later Telia). Over the next three years, five other countries introduced the service: Denmark, Finland, Norway, Spain and Austria. The first UK mobile-phone network, Racal’s Vodafone, went live on 1 January 1985. Eight days later, its competitor, BT’s Cellnet (now O2), was launched.

Vodafone would go on to transform the world’s telecoms market, but few could have imagined that in the mid-1980s. It emerged from an obscure British company, Racal Electronics (its name derives from those of its two founders, Ray Brown and Calder Cunningham), which specialised in defence equipment and in particular military radio communications. In the mid- to late 1970s its profits grew dramatically and it looked unstoppable. But by 1982 the profit progression was beginning to go into reverse.

During 1982, when the first Thatcher government in the UK announced plans to issue two mobile-phone licences, Racal bid for the second. Its dynamic chief executive Ernie Harrison thought the new service would provide the company with an opportunity to sell more radios – not military radios, but commercial cellular radios, or mobile phones. Racal set up a joint venture with US company Millicom and Hambros Advanced Technology Trust, a fund that had been established by the venerable Hambros Bank.

Bidding against a consortium led by Cable & Wireless – formed in 1929 and until the previous year owned by the British state – and a team that included Ferranti (the company responsible for the world’s first commercial computer) and London Car Telephones (which had been an agent for BT’s pre-cellular car-telephone service), Racal looked to be outgunned. However, on the day of the pitch, it managed to strike all the right notes, and Racal Millicom was duly awarded the country’s second licence in 1983.

Racal Millicom now had just under two years in which to design, build and launch its network. It selected Ericsson’s AXE switch, on the grounds that it was the biggest and most powerful system available at the time. It adopted a highly pragmatic approach to site locations, making use of anywhere that came to hand. And it looked for people to run the new business. Gerry Whent was to head the new outfit, with Chris Gent as the managing director of the network business and Julian Horn-Smith to head the marketing team and, later, international development.

Saatchi and Saatchi advertising agency was called in and asked to come up with a name. This resulted in ‘Vodafone’, a portmanteau of ‘voice’ and ‘data(f)one’. The new company located itself in the small country town of Newbury in Berkshire some 80 kilometres (50 miles) west of London.

‘Hi Dad. It’s Mike. This is the first-ever call made on a UK commercial mobile network.’ This is how the first call made over the Vodafone network began, just after midnight on 1 January 1985. Michael Harrison, one of the CEO’s sons, had gone to Trafalgar Square to join in the New Year’s celebrations armed with a Vodafone VT1 mobile handset, and he’d used it to call his father. Vodafone was up and running a full three months before the date stipulated in its licence – and, just as important, eight days before Cellnet, its competitor.

The handset Michael Harrison had used weighed something in the region of five kilograms and it wasn’t cheap – but then, nor were any mobile phones back then. Vodafone’s basic car telephone sold for £1,415 and top-end models could cost anything up to £3,000 – about £8,000 in today’s money.

Indeed, one of the key decisions facing the two new companies was how to price the new service. BT’s existing phone tariffs were complex, with call charges varying according to distance and time of day, and depending on whether the call was made during peak, standard or off-peak hours. Cellnet, it seemed, planned to replicate this approach within the new industry.

Whent, however, was determined to keep the business simple: Vodafone’s proposed charges were a £50 fee to join the network, a monthly subscription charge of £25 and a per-minute call charge of 25p. The simplicity of the tariff appealed to dealers and customers alike, and a few days later Cellnet announced an almost identical package.

By the end of the 1984/85 financial year three months later, Racal announced that it had connected 2,711 subscribers. This didn’t match its competitor’s figure of 4,500, but then the Cellnet’s numbers were augmented by a steady flow of customers from BT’s pre-cellular car-phone services.

As the weeks passed, the numbers on the new networks continued to increase. By the end of the two companies’ 1985/86 financial year, Vodafone had narrowed the gap between the two networks to just 3,000. It drew level with Cellnet three months later, and in July 1986 it took the lead. (It would be nearly 20 years before Racal would drop behind Cellnet again, in December 2005, by which time the number of subscribers on the network had become rather less important than the profits they generated: the two company’s UK operating profits for that year were £975 million for Vodafone, and £623 million for Cellnet.)

There were several factors behind Vodafone’s victory in this all-important early phase. First, there’s no doubt that it benefited from BT’s poor image. Although it had been part-privatised, BT was still at heart an old-school monopoly, with all that that implies. Cellnet itself was different, and better than that, but the public weren’t dealing with Cellnet, they were dealing with its service provider, British Telecom Mobile Communications – so Cellnet became tarred with the BT brush.

A second factor was the two business’s use of these independent service providers. Vodafone understood that they were there to make money, so the best way to treat them was to ensure that they understood the intricacies of service and were provided with all the necessary support. Cellnet preferred to hand out expensive freebies instead.

Third, there was the issue of network quality. Cellnet had opted for a highly scalable, distributed architecture based around Motorola switches, with radio links engineered in-house by BT, but in practice the system didn’t live up to expectations. And despite access to BT’s vast property portfolio, Cellnet’s coverage wasn’t as extensive as Vodafone’s – Cellnet had located its base stations inside or on top of local telephone exchanges even if those weren’t the best places to put them.

Vodafone deserved its lead – it had been more flexible, more responsive and more imaginative. By the end of 1986 the company was connecting over 1,000 customers a week; it had more than 63,000 subscribers, a lead of 9,000 over Cellnet. And it had moved into profit on a day-to-day basis some months ahead of expectations.

However, in October 1987 stock markets around the world began to plummet. In just a few weeks, Racal’s shares fell from an early October high of 350p to about 200p. In the months that followed, Racal’s management watched the company’s share price slowly creep up again to around 240p. This was worrying: at that level, the Racal Group was valued at just £1.5 billion. They believed Vodafone alone was worth at least that much.

The share price was important because a higher share price made it easier to raise additional funds; the other side of this coin was that a low share price, which undervalued a business, left the company vulnerable to a possible hostile takeover, as Racal was soon to discover.

The rumours that Eric Sharp of Cable & Wireless had been asking questions about Racal, and in particular the Vodafone business, were confirmed by a friendly broker who’d spotted some atypical trading in Racal shares. It took Cable & Wireless just two days to acquire a 2.8% stake in Racal – just below the 3% threshold, the point at which it would have been required by law to make its holding public.

By this time, Harrison was well aware of Cable & Wireless’s intentions. He knew he had to raise the company’s value significantly if it were to stand any chance of defeating a hostile bid, but Racal’s stock was already trading at a premium and many investors remained sceptical. So he decided to unlock the value within Vodafone by spinning it off as a separate company.

In 1988 Racal announced that it intended to float 20% of Racal Telecom on the London and New York stock exchanges. Racal’s shares rocketed, gaining 66p that afternoon before climbing to a new high two days later. It was announced that Racal Telecom was to be valued at £1.7 billion, 170p per share – way above the market average.

Investors in the UK remained lukewarm, but US fund managers took a different view and went on to buy a majority of the 200 million shares that were available, giving Racal Telecom quite a high profile on Wall Street.

The threat from Cable & Wireless had been averted, at least for a little while.

In late 1989 the UK’s Department of Trade and Industry stated that it intended to increase competition in the mobile market through the offer of three new licences. The new entrants wouldn’t be operating ‘old-fashioned’ cellular-telephone systems, but personal communications networks (PCNs), using an exciting new technology called digital communication system 1800 (DCS-1800), which was later referred to as GSM-1800 (the number refers to the operating frequency).

The licences went to Microtel (of which Hutchison Telecom, a subsidiary of the Hong Kong conglomerate Hutchison Whampoa, ended up owning 65%), the Cable & Wireless–Motorola joint venture Mercury PCN, and Unitel. The latter two merged in 1992 to create Mercury one2one.

In the meantime Racal Telecom continued to run rings round Cellnet, in both market share and profitability: it announced annual profits of £165 million for its 1989/90 financial year while its competitor was still barely profitable. Yet once again its share price began to languish: the state of the UK economy wasn’t great and the news that the government was to award three new mobile network licences didn’t help.

In mid-1991, although it had its own mobile business by now, Cable & Wireless came back to try again. This time it had brought some extra muscle in the shape of General Electric Company (GEC), Britain’s largest electronics company. Harrison immediately summoned the board for an emergency meeting, during which he suggested that the Racal Telecom business was now big enough to stand on its own two feet and that, were it to be cut loose, the danger of a hostile takeover of both Racal companies – Racal Electronics and Racal Telecom – could be averted.

The next day plans for a complete demerger of Racal Telecom were posted with the London Stock Exchange. This had the desired effect and the rising prices of both Racal stocks put the business out of reach. Once again a predator’s threat had been avoided.

Racal Electronics distributed its 800 million shares in Racal Telecom to its shareholders in September 1991 and the former subsidiary formally adopted a new name, Vodafone Group plc. It became a constituent of the London Stock Exchange’s FTSE-100 index, meaning there was a huge immediate demand for the shares.

In 1993 the Mercury one2one network eventually went live, but to the surprise of many, the service was available only in the area inside the M25, the orbital motorway that circles London. According to one2one, the geographic limitation wasn’t that important, as most calls from mobiles take place within a few kilometres of home, and the area inside the road accounted for over one third of all UK calls. But the public didn’t buy the limited-coverage argument: people wanted a phone that would work anywhere.

In the meantime, Hutchison had taken its time building its network to ensure that it could offer something close to national coverage when it eventually launched. This it did in 1994, Microtel becoming Orange and launching the world’s first national DCS (GSM)-1800 network. An Orange subscription came with a bundle of ‘free’ minutes (they weren’t actually free because a user couldn’t take the minutes on their own without a subscription), and the service included voicemail, call-waiting, caller identification and many other new features. Importantly, calls were billed by the second, not rounded up to the next full minute. Subscribers could choose an 1800MHz variant of Nokia’s new 2110 handset.

Orange’s biggest achievement was that it managed to persuade people that mobile phones weren’t just for the businessman or the more fortunate – they were for everyone.

Many of the governments of Europe took the launch of the new digital GSM networks in the early 1990s as an opportunity to introduce competition into their markets. One after another, administrations across Europe announced plans for second licences. This policy was subsequently adopted in much of the rest of the world. Vodafone and its American rival AirTouch were the first to appreciate the significance of this development, and together the two led the charge to obtain these valuable licences.

One of the first and biggest opportunities arose in Germany, which was an undeveloped market, with just 99,900 mobile connections at the end of 1988 (a penetration rate of just 0.13%). (The UK, by contrast, had nearly five times that many and a penetration of 0.9%.) Various consortia were created and seven bids submitted. As 1990 wore on, it became clear that two were some way ahead of the pack: BMW, which had brought in Veba AG, a German electrical utility, and Vodafone, to provide the industry expertise; and Mannesmann, with PacTel and a local bank. The Mannesmann–PacTel team was awarded the licence.

In 1992 Vodafone won one of the two available licences in Greece, taking a 45% stake in the company that would become known as Panafon. This was to prove a great success over the next few years.

Then, in the same year, a second opportunity arose in Germany. The third digital network (after Deutsche Telekom’s D1 and the new operator, Mannesmann’s Mobilfunk) would use the same DCS-1800/GSM-1800 technology employed by the PCN operators in the UK. A number of the companies that had been unsuccessful in the 1990 contest teamed up to bid for the new licence. BMW was one, and this time it recruited Hutchison and GTE; and BellSouth teamed with Thyssen, a heavy-engineering company based in Dusseldorf. Julian Horn-Smith roped in Veba, the energy company that had been part of BMW’s original consortium and, together, the two edged their way into BellSouth’s E-Plus consortium – which got the nod.

Vodafone was unlucky in Italy, however. In 1993, under the name Unitel, it put together a strong team including the giant car company Fiat, and Fininvest, an investment company controlled by media tycoon and politician Silvio Berlusconi. It faced competition from four other groups. Olivetti, the Italian typewriter-to-technology conglomerate, had joined forces with Bell Atlantic to form the Omnitel consortium, and that looked to be the biggest threat. Unitel was cruising towards victory but then Berlusconi announced that he was entering politics, heading a new political party. To the amazement of almost all the pundits he won a snap general election – on the same day the licence was due to be awarded. To avoid the obvious charge of nepotism, it had to go to the Omnitel team.

In Australia the state-owned monopoly held the first licence, while a team headed by Cable & Wireless and including BellSouth, which had bought Australia’s international satellite company, was given the second mobile licence as part of the deal. Vodafone’s 95%-owned Australian subsidiary bought the third licence.

Another opportunity arose, in South Africa in 1993. Ahead of the election of the new democratic government, representatives of Nelson Mandela’s party decided to open up the telecoms market by awarding two new GSM licences. The incoming administration wanted an international partner to help Telkom, the national PTT, operate the new GSM system, and Vodafone was selected from the shortlist. A new company, Vodacom, was formed, in which Vodafone took a 35% interest. (The second licence went to Mobile Telephone Networks Holdings, or MTN, a consortium that included Cable & Wireless.)

In 1999, Vodafone acquired AirTouch. This gave it by far the largest portfolio of international mobile assets, including a 34.8% interest in Mannesmann Mobilfunk, with Mannesmann holding the balance, and 21.8% of Omnitel Pronto Italia, where Mannesmann had just raised its stake to over 55%. Following Mannesmann’s acquisition of Orange in the UK, Vodafone obtained control of Mannesmann – the first time a major German company had fallen to a hostile bid, and the largest-ever takeover in corporate history. The two deals made the UK company the largest mobile operator in the world – less than two decades after its formation.

Prior to the collapse of the Berlin Wall in 1989 and the subsequent collapse of the Soviet Union, even the best telephone networks in Eastern Europe suffered from decades of under-investment. In many countries there were large numbers of towns or villages with no phone service at all.

East Germany was typical. When Deutsche Telekom took on the responsibility of providing modern communications to the 16 million new German citizens in the New Federal States (those in the former East Germany that were reincorporated into the Federal Republic of Germany on reunification in October 1990), it discovered that there were more than 2,000 communities without a single phone, and that most of the switching and transmission plant dated back to the late 1930s.

All of the other countries were in a similarly sad position. The Hungarian telephone company Magyar Telekom (Matav) was one of the few that admitted it. Its 1988 Annual Report included a set of statistics that showed that in every respect bar one, it fell well below the standards it wanted to meet, with far lower overall phone ownership, a network that experienced more line faults per month than a modern operator experienced in a century, and waiting lists of seven years in Budapest and a shocking 14 years elsewhere.

The ending of communist rule gave rise to many investment opportunities in the newly liberated countries, although at first there wasn’t a huge amount of enthusiasm. US West was one of the first to become active in the region, when it applied for several of the new regional cellular licences in Russia. By 1995 it had stakes in cellular businesses in Hungary, the Czech and Slovak republics, and, of course, Russia. These holdings were complemented by several key investments in Western Europe, including a 50% stake in the UK’s Mercury one2one, and a 26.8% interest in TeleWest, one of the largest of the UK’s new regional cable TV companies.

US West chose its partners pretty well. In the UK it had teamed up with Cable & Wireless and TCI, the largest US CATV company, while Bell Atlantic was selected to be its partner in the Czech and Slovak republics. In Hungary, it went with the local phone company Matav, which would itself soon come under the control of a joint venture between Ameritech and Deutsche Telekom, the company that was to invest the most in the revival of the phone systems in these countries.

Apart from its considerable investment in Hungary, Ameritech’s interests were confined to an equal share of a 49% stake in Telekomunikacja Polska, Poland’s national phone company. France Telecom was Ameritech’s partner in this venture and would eventually buy out its partner. This, at least for the moment, was France Telecom’s only investment in Eastern Europe.

Deutsche Telekom, by contrast, had looked to establish a strong presence in the region almost immediately. In 1992 it became one of three foreign investors in Ukraine’s first cellular operator, an NMT-450 business called Ukrainian Mobile Communications. In 1993 it became an early investor in Russian cellular, taking an initial 23.5% shareholding in Mobile TeleSystems (MTS), a company that would become one of that country’s big three cellular operators. In the same year, it entered the joint venture with Ameritech in Hungary, acquiring a 30% interest in Matav. Investments in the second cellular companies in Poland, the Czech Republic and Slovakia were to follow over the next two to three years, then a 35% stake in Croatia’s Hrvatski Telecom.

Later, its Hungarian subsidiary Matav was also encouraged to expand in the region. It acquired a 51% stake in an entity that controlled Macedonian Telecom in 2001, following that in 2005 with 51% of Telekom Montenegro.

The second licences in Eastern Europe gave some of the smaller European PTTs a chance to expand their footprints. One of the best opportunities arose in 1993, when Telecom Finland (later Sonera Corporation), together with the Cukurova Group, Ericsson and some local investors, established Turkey’s first GSM operator, Turkcell. This astute move provided Sonera with a springboard for further investment in Moldova, Georgia, Azerbaijan and Kazakhstan.

The following year Telenor, Telia and the Dutch PTT Telecom (later Koninklijke KPN or Royal KPN) all got in on the act, becoming partners with Sonera in the second mobile licence in Hungary. Telia and Sonera worked in partnership in the Baltic region, acquiring control of mobile companies and their PTT parents in Latvia, Lithuania and Estonia between 1994 and 1998. Telenor, meanwhile, acquired companies in Montenegro, Ukraine, Serbia and Bulgaria between 1996 and 2013.

The Eastern European licences have generally proved to be good investments, but looking back, it seems extraordinary that Europe’s leading telecoms operators showed so little interest in the licences that were on offer in Western Europe. By the time Europe’s giant PTTs had woken up to the value of cellular, all of the most valuable had gone. AirTouch had bagged by far the largest hoard, with Vodafone some way behind in second place. France Telecom had woken up in time to get the second licence in Belgium and Deutsche Telekom, the second in Austria; Telecom Italia won the other Greek licence (awarded at the same time as Vodafone’s), but BT and Telefónica had nothing. Nor did any of the smaller players.

Why were the PTTs so unconvinced by the new service, at least initially? The management in all of Europe’s PTTs – indeed, of every incumbent phone company in the world – had grown up believing in the value of the fixed networks they were building, and they couldn’t bring themselves to believe that that could possibly change. So they persuaded themselves that mobile would only ever be a marginal thing, because they couldn’t cope with the idea that it might ever be more than that.

There was plenty of evidence to the contrary, however, and the market was growing daily. Digital technology had arrived and was maturing; the price of handsets was declining; and prepaid-billing arrangements (pay-as-you-go) had been launched.

Telecomunicações Móveis Nacionais (TMN), the mobile arm of Portugal Telecom, was the first company to launch a prepaid tariff in Europe, and the effect on the business of the new tariffs was dramatic. The company had ended 1994 with 85,000 mobile connections; a year later, the number was 171,000; 1996 saw an increase to 332,000; and 1997 that number leap to 762,000. During 1998 TMN’s total customer base reached 1.43 million, over 80% of which used prepaid billing.

Within a couple of years, almost every other operator in Europe had followed TMN’s lead. After a decade of mostly monopoly operation (from 1981 to 1991), Europe had a total of 4.4 million mobile subscribers. By the end of 1995, the total was up to 23.7 million, and by the end of 1999, 63 million people connected through prepaid tariffs and a further 112 million on contracts.

Today, although the USA still lags in prepaid numbers with fewer than 15% of all connections billed through this method, around 55% of all European connections are prepaid. In Asia the ratio is over 75%, despite the almost total absence of prepaid in Japan. Latin America is 90% prepaid and Africa is over 95%. The global average at the end of 2018 was around 75%, or about 5.9 billion people.

The switch from analogue to digital

From the birth of the first European mobile networks in the 1980s, the industry was on the search for a digital standard that would solve the problem of incompatibility between networks in neighbouring countries. The European Conference of Postal and Telecommunications Administrations had created a study group, Groupe Speciale Mobile (GSM or Special Mobile Group; GSM later came to stand for ‘global system for mobile’), in 1982 to work towards a common European standard based on digital technology. The leading phone companies of Europe nominated representatives to the GSM.

By 1987 the GSM had agreed to focus on a technique called time-division multiple access, or TDMA, one of several possible methods of multiplexing separate signals. The transmission of multiple signals is like a room where numerous people want to talk to each other at the same time, and one way to make this possible is for individual speakers to be disciplined and only talk in turn – this is essentially time division. (Another option is to use different pitches, called frequency division, while a third, more radical solution, is to speak in different languages, called code division, so that those in the room who understand the language will listen while the rest will tune out.)

By 1989 oversight of the project had passed to a body called the European Technology Standards Institute, and by 1991 the first networks based on the new GSM standard had started operating. One key element of the new GSM standard was that the identity of the user wasn’t defined by the handset itself, but rather by a ‘subscriber identity module’ or SIM. This allowed subscribers to use more than one device, switching the SIM from, for example, a hand-portable phone to a car phone or a data modem. Later it would enable customers of one network to leave and join another without the need to buy a new handset.

Despite the standard’s many advantages, initial uptake was slow. There were several reasons for this. The new GSM handsets were much more expensive than the latest analogue designs, and a lot larger. And coverage didn’t match that of the existing cellular systems – in more than a few networks, the ‘Swiss cheese’ effect (‘holes’ in the network) disrupted the best efforts of the network planning teams.

And there was another problem: the GSM solution wasn’t alone in the market. Digital AMPS (D-AMPS) was a fairly straightforward adaptation of the US AMPS standard using the existing AMPS channels; Personal Digital Cellular (PDC) was similar to both GSM and D-AMPS, in that it was also a TDMA-based standard, but it didn’t work very well unless the radio signal was very strong; and the proponents of Code-Division Multiple Access (CDMA) maintained that it was superior to all TDMA technologies. The problem was, the technology wasn’t quite ready to be deployed.

By the time the first GSM networks in Europe had been running for a few months, many operators in other parts of the world had seen enough to decide in favour of the European standard. By 1995 the European technology had been adopted as the standard by countries with comfortably more than 50% of the world’s population.