South African Telecommunications: History and Prospects

[rpm]

Robert B. Horwitz

1.0 Introduction

Covering the broad tip of southern Africa and possessed of vast mineral deposits, South Africa is the most important industrial nation in Africa. Historically, it has been among the world's largest gold and diamond producers and continues to derive at le ast one-seventh of its gross domestic product (GDP) from mining and quarrying. At R431.5 billion in 1994, South Africa's GDP is by far the largest in Africa. 1

Of course, what is unique about South Africa is its politics. For the most part, white domination of the country s majority black population has been the rule since Dutch settlers arrived in the Cape peninsula in the latter half of the seventeenth centur y. Although serious hostility has always existed between Afrikaners (descendants of mostly Dutch settlers) and the English (who began settling the territory after Britain seized the Cape in 1806 to protect its Indian sea route), they essentially made comm on cause when it came to the use of black labor. This was particularly true after diamonds and gold were discovered in the late 1800s. The emergence of South Africa as an independent republic in 1910 rested on policies explicitly designed to ensure that w hites retained political power and control over the state. A wide range of measures ended the limited access of "Coloureds" (mixed race) and Africans to the vote and required that all senior positions in the state bureaucracy be filled by whites . Historically, the South African republic has been, quite simply, a democracy for whites. 2

In 1948, South Africa s system of racial separation became dramatically explicit and manifested itself in a series of nefarious laws known as "Apartheid." Two central and essentially contradictory characteristics, however, were prominent. First , apartheid aimed to control black labor. The Land Acts of 1913 and 1936, for example, pushed blacks off productive farmland not only to safeguard such land for whites but to compel blacks to seek menial jobs in mines and factories. The so-called civilize d labor policy entailed paying whites at a higher rate than blacks for doing unskilled or semiskilled jobs, and in most economic sectors a color bar kept blacks from skilled occupations. Second, apartheid was constructed ostensibly on the premise of inher ent differences between races and nations--and the desirability of their separation. Hence, South Africans were formally classified according to race (white, mixed race, or "Coloured," Indian or "Asian," black or "African") a nd tribe (usually by black language grouping), and South African citizens were required by law to live in the particular geographic areas set aside for them. Indeed, since the formal rise to power of Afrikaner nationalism in 1948, the ruling whites stated goal had been to secure a South Africa with no black citizens. Blacks were to reside only in their homelands, supposedly under their own "tribal" political structures and with responsibility for their own separate institutions and economic deve lopment.

Together, these two apartheid policies succeeded in securing a low-wage industrial reserve army for South Africa but created at the same time vast inefficiencies. Because black "guest workers," for example, were needed in cities but were forced to live far outside them, South Africa s transport system had to accommodate massive long-range daily population movements--at the cost of considerable state-provided subsidies. Outlays for police and security represented other inefficiencies: while alwa ys high, these outlays reached immense proportions in the years following the 1976 uprising in Soweto.

In 1993, the population of South Africa, including the ten black homelands, was approximately 38.5 million. Blacks made up 75.6 percent of this total, Asians (primarily people of Indian ancestry) 2.6 percent, mixed race inhabitants 8.6 percent, and white s 13.2 percent. In the mid-1990s, some 89 percent of Asians, mixed race South Africans, and whites were urbanized, compared to only 50 percent of the country s Africans. In 1992, more than half the African population was under nineteen years of age (South Africa Institute, 2). The decades of legalized discrimination and violence against Africans resulted in a remarkably unequal society. Employed Africans, for example, are situated overwhelmingly in semiskilled and unskilled labor markets, and although acc urate statistics are often difficult to obtain, unemployment among Africans in the early 1990s was estimated at 45 to 46 percent of the economically active population (South Africa Institute 1992, xliii; Business Day , August 23, 1993. The repeal of "influx control" laws in 1986 which brought about a 28 percent increase in the number of urbanized blacks did not, however, end the nation s economic inequality. While the ratio of white-to-African real earnings in South Africa s manufacturing sector has been converging in recent years, in 1989 it was still approximately 3.4 to 1 (South Africa Institute 1992, 259). Moreover, a sizable percentage of Africans lives in informal settlements and squatter camps situate d at the edges of black townships, without access to tap water or sewerage facilities, much less electricity or telephones. 3 Even with the formal end of the apartheid state per capita expenditure on schools finds Africans receiving just 25 percent of traditionally white school expenditures (South Africa Institute 1992, 195). Unemployment levels remained unchanged and proved to be one of the most intractible problems facing the popularly elected government elected in 1994.

It is in politics that the major dramatic changes in South Africa have occurred. The concatenation of three forces--widespread overt opposition to racial rule, efforts of the state to preserve white minority control, and international pressure--gave rise to a process of debilitating economic crisis and intensifying political conflict that placed intense pressure on the South African state (Price 1991). In February 1990, the government of F. W. de Klerk legalized banned organizations, including the Africa n National Congress (ANC) and the South African Communist Party (SACP), and began the process of rescinding the formal laws of apartheid. 4 This led to tremendous political activity on all fronts, a great deal of conflict (some violent), and a general worsening of the economy. In the period after the legalization of the ANC, representatives of the black majority engaged in difficult negotiati ons with the white government to establish new political ground rules and structures, including the resolution of fundamental questions over power-sharing arrangements versus one-person, one-vote democracy. Struggles within the African population itself a lso arose, the most prominent being the frequently violent confrontations between supporters of the ANC and those of the Inkatha Freedom Party (see Adam and Moodley 1992). Momentous elections held in April 1994 brought to power a coalition dominated by th e ANC, which called itself a the "Government Of National Unity (GNU)."

Less noticed than these highly visible political changes was an important transformation in South Africa's political economy. Long a strongly interventionist state characterized by extensive government involvement in the economy, in the recent years of e conomic crisis, the South African state began to dramatically scale back its economic activity. Historically, much of the state's intervention in the economy involved controlling labor markets under the apartheid system, and apartheid secured a continuous flow of phenomenally cheap black labor to white farms, mines, and factories. But state intervention in South Africa also entailed the operation of monopolistic public corporations--known as "parastatals"--in many areas of economic life. In the past few years , the South African government began disengaging from its extensive historic involvement in the parastatals. Telecommunications is one of several sectors--including electricity and transport, oil, and iron and steel production-- where this retrenchment of state intervention took place.

 

[rpm]

2.0 THE "ANCIENT REGIME"

Historically, the South African Posts & Telecommunications (SAPT) was in most respects a classic post, telephone, and telegraph (PTT) monopoly, legally monopolizing postal and telecommunications services and operating a system characterized by intern al cross-subsidies (Statutes of the Republic of South Africa 1958). Perhaps the clearest cross-subsidy was that from telecommunications to posts. With the entire operation statutorily forbidden either to make a profit or experience loss, profitable teleco mmunications services balanced out large operating losses on the postal side. The SAPT, often colloquially referred to simply as "the Post Office," was classified as a "state business enterprise" and was run through the office of the M inister of Transport and Communications. 5 After the emergence of the South African Republic (most often referred to as "Union") in 1910, the finances of the Post Office were controlled by the Treasury. Post Office revenue was paid into the Exchequer and all Post Office expenditures came from the Exchequer. Because it operated directly out of a government ministry and hence was tied to Parliament's annual planning cycle, the SAPT generally experienced close financial oversight. The oversight system evolved such that SAPT had to submit de tailed annual reports on its budgets, accounts, and proposed capital expenditures to the Auditor-General (an instrument of Parliament), who then wrote a report on the SAPT.

Under this system of oversight, the SAPT as a rule experienced little outright corruption or gross overspending but generally suffered from a shortage of capital inasmuch as it had to compete for funds with other central government capital projects. Clos e government oversight also meant that SAPT did not set its tariffs according to marginal costing principles. Since members of South Africa s Parliament felt that they had to serve their constituents' interests in low telephone rates, tariffs were set wel l below the level that would have enabled the SAPT to meet demand expediently. If upward tariffs were thought to contribute to inflation or proved politically difficult for the transport and communications minister, they were typically adjusted downward ( Taylor 1992). In 1968, South Africa s system of financing changed to some degree when the Post Office's finances were separated by law from the Exchequer and placed under its own control. Yet while SAPT controlled its revenues, it was still dependent on t reasury loans to finance capital expenditures. As in many nations, the SAPT operated a savings bank that provided for some of its loan requirements. The operation of the bank itself, however, was an expensive undertaking. It was only in 1972 that SAPT fin ally received permission to seek its own local or offshore financing (Financial Mail, October 3, 1986, 47-48).

Judging by comparative productivity indexes, one could reasonably conclude that the SAPT functioned as a repository for public employment. For instance, on the telecommunications side access lines in service per employee measured a relatively low 45.4 as late as 1989. By way of contrast, South Korea measured 226.3, the United States 130.6, and Mexico 95.6. Even Turkey outperformed South Africa at 65.9 access lines in service per employee, according to 1989 data (Coopers & Lybrand 1992, 16). 6 This performance, however, must be understood in the broad context of South Africa s apartheid variant of the interventionist state: South African parastatals historically have been the focus of the "job reservation" system, whereby unemployable whites (primarily poor Afrikaners who had left farming) were given jobs by the state or its parastatals. At the very least, the SAPT participated in the job reservation system indirectly, through the contracts the Post Office fashioned with domestic equi pment manufacturing companies, which particularly in the early years were under mandate to hire whites (Kaplan 1990, 31-32). 7

2.1 SAPT and Local Telecommunications Equipment Manufacturing

The relationship between SAPT and domestic equipment providers functioned, as in other nations, in the manner of an industrial policy, with local purchasing obligations and high domestic content requirements. In the main, the policy succeeded rather well . By the mid-1990s, the manufacture of telecommunications equipment constituted the largest part of the South African electronics sector. Prior to 1958, SAPT's equipment was supplied by the principal contractors to the British Post Office: Automatic Telep hone Electric Company, Siemens U.K., and Standard Telephones and Cables. In 1958, with military and strategic considerations in mind, the government utilized the SAPT s monopsonistic power to establish extensive local production of telecommunications equi pment. Long-term contracts (officially known as the Manufacture and Supply Agreements) with five principal South African telecommunications equipment companies embodied national goals on local content (Kaplan 1990, 27-31, 79-85).

In the mid-1990s, South Africa s four remaining major telecommunications equipment companies were the Altech Group (STC and Teltech), Siemens, Telephone Manufacturers of South Africa (Temsa), and Plessey, but other companies, including General Electric C ompany (GEC) and ATC, have become significant players in local equipment provision as well. Altech was South Africa s exclusive supplier of transmission equipment while Siemens, Altech, and Temsa provided switches. Siemens also supplied telex and teletex equipment as well as telephone exchange power supply equipment, and Temsa, originally set up by the local Plessey and GEC companies, was the sole manufacturer of standard telephones. Plessey was responsible for providing small private branch exchanges (PB Xs), and Aberdare and ATC were selected to produce cable for South Africa s national transmission network. In the early 1990s, ATC was the only manufacturer of fiber-optic cable in South Africa. More recently, Telkor was awarded the contract to supply Sou th Africa s pay phones (Telkom 1991, 31).

All the principal local producers of South Africa s telecommunications equipment and components have a significant part of their equity held by one of South Africa s largest local corporate groups. 8 Anglo-American, for example, has a 20 percent stake in Altech, and the Sanlam Group has a 32 percent stake in Siemens. Considerable ownership changes in South Africa s telecommunications manufacturing companies occurred in the 1990s, reflecting a general rationalization and a specific regrouping in the advent of a sharp decline in orders from Telkom, South Africa s telecommunications entity since 1991 (see Kaplan, this volume, chapter ?). GEC now holds a two-thirds stake in Temsa with the other third held by Siemens. ATC is part of the Reunert Group, owned (80 percent) by Barlow Rand, and Reunert is also the owner of Telkor. South Africa s principal telecommunications companies all have a shareholding in South African Micro Electronic Systems (SAMES), Sou th Africa s sole local producer of integrated circuits (Kaplan 1990, 84-85). Following a major fire in 1990, SAMES was restructured, with the Industrial Development Corporation, a state agency, taking a majority stake.

Historically, as technology became more complex the local South African companies essentially became licensees of First World companies. While in many respects these foreign corporate telecommunications connections follow broader historical relationships (particularly with the United Kingdom), the SAPT's decision in the late 1970s to digitalize the network tightened the connections with the foreign equipment suppliers. The decision was based on the fact that demand was increasing and the British-based el ectromechanical switching system had begun to present SAPT with serious technical problems. However, small nations cannot easily amass the capital and expertise to create and manufacture highly advanced, general (as opposed to niche) telecommunications de vices. As a result, South African equipment companies entered into licensing arrangements to assemble foreign-designed telecommunications equipment in South Africa. For example, Siemens licensed Temsa to manufacture the German EWSD digital exchange, and T eltech (part of the Altech group) was licensed to manufacture the French Alcatel E-10 digital exchange (designated locally as the SA 128E). Inasmuch as the late 1970s represented the early days of digital technology, the choice of suppliers was limited, a nd only a few companies could present viable digital equipment offerings at that time.

 

[rpm]

The decision to authorize two different foreign telecommunications equipment suppliers for South Africa was a political one. Given economic sanctions and the possibility of a supplier cutoff, dependence on a single source of foreign technology was consid ered to entail political risks. The government thought that the difficulties of meshing two different digital exchange systems was outweighed by the political slack offered by two different foreign suppliers (Kaplan 1990, 41-42). In the mid-1990s, foreign owners thus have significant percentages of shareholding in most of the major South African telecommunications companies. For years, for example, Alcatel (France) had a 12.5 percent stake in Altech. In January 1993, Alcatel boosted its stake to 50 percen t and allowed Altech to buy a stake in Alcatel as well, thus more fully integrating the companies. In the mid-1990s, Siemens (Zurich) had a 52 percent ownership of Siemens South Africa, and until the recent merger of Plessey and Grinaker, Plessey (United Kingdom) had a 74 percent stake in Plessey South Africa. Following the reorganization of Plessey, GEC of the United Kingdom now holds 50 percent of GEC South Africa's stake in Temsa.

Originally, South Africa s Manufacture and Supply Agreements covered a ten-year term, but the equipment suppliers persuaded the SAPT to increase the term to fifteen years because of the high investment in plant and technological transfer from overseas pr incipals. 9 SAPT was obliged to purchase its equipment from the domestic companies when feasible, depending upon the percentage of local content and the comparison of local with international equipment prices. According to policy, SAPT would generally purchase South African-made equipment if there was a reasonable degree of local content so long as the price did not exceed the international price by more than 25 percent. In many instances, however, the price premium was far higher, ranging from 32 percent for telecom munications equipment and 60 percent for electronic components (Kaplan 1990, 106-26). 10

While the Manufacture and Supply Agreements worked well to ensconce viable equipment firms and establish coherent standards and specifications, the pricing structures may have acted to replicate the famous Averch-Johnson-Welliscz effect. As a rule, SAPT would take back a percentage of high supplier margins as a kind of rebate or "profit sharing." Designed to give suppliers incentives to enhance efficiency, the rebate arrangement instead gave suppliers incentives to reinvest in their own plants (such investment being part of the overall cost structure) in order to reduce the amount that SAPT would take back (Hartyani 1992; Schulze 1992; Averch and Johnson 1962; Welliscz 1963). Although intended to foster innovation and product development, South Africa s supply agreements paradoxically may have hindered them. According to David Kaplan (1990, 99-101), the agreements function to exclude the entry of new smaller companies that might be highly innovative. The agreements provide alternative and less risky routes to achieving high levels of profitability. 11 As if to underscore the historical weakness of South African telecommunications equipment manufacturing, the South African Board of Trade and Industry reported that telecommunications manufacturers exported only 1.5 percent of their products in 1985 (Kapl an 1992). Since then, however, local telecommunications equipment exports have improved. Exports, primarily to Eastern Europe and Africa, expanded after 1988, increasing two and one half times in constant Rands (see Kaplan, this volume, chapter ?). 12 By 1991, the telecommunications equipment industry in South Africa was worth R1.839 billion (BMI TechKnowledge 1992, 95).

 

[rpm]

3.0 A NEW TELECOMMUNICATIONS REGIME

South Africa s telecommunications structure changed in October 1991 when posts and telecommunications were separated from each other and set free from direct ministerial control (Republic of South Africa 1991c). A combination of forces, including those t hat have affected other African telecommunications administrations as well as those specific to South African politics, led to the regime change (see Horwitz 1992). Demands from large users--who had formed effective lobbying groups--for improved service q uality, freedom of choice, and the opportunity to compete began to find a receptive audience in the government. Similarly, on the technological front, rapid advances, particularly in the areas of fiber-optic and mobile communication, had begun to erode th e telecommunications sector s natural monopoly conditions. Finally, and perhaps most important, in the rapidly declining economic climate of the 1980s, the government came to the tradition-jolting conclusion that privatizing its public corporations and en terprises was a better option than operating debt-ridden ones. This historic reversal was the result of a series of official investigations of the principal South African parastatals, including the SAPT, conducted by Dr. W. J. de Villiers and known as the De Villiers Report .

3.1 Telecommunications Reform: The De Villiers Report

In the case of the SAPT, de Villiers found that South Africa s telecommunications system was characterized by very heavy debt--a consequence, he concluded, of overforecast demand and "goldplating," which he characterized as a too rapid conversio n to digital technology (De Villiers 1989). SAPT's staggering debt stemmed from the timing of its expansion. With the telephone market for whites saturated, the SAPT began in the late 1970s to extend service to Indian and "Coloured" areas and, l ater, to a number of black areas. In the midst of this expansion, SAPT faced the problem of whether to expand its problematic electromechanical switching system or go digital. It decided to pursue the digital option. These expansion projects required a va st infusion of capital, which the SAPT was permitted to obtain through loans on the international capital market. The annual capital expenditure on telecommunications equipment increased fivefold over the seven years from 1980-81 to 1986-87 (De Villiers 1 989, concise version, 11). As South Africa's economic condition deteriorated in the 1980s and the value of the Rand plummeted, the SAPT's--and the government's--indebtedness increased dramatically. Similar stories characterized the state of South Africa s electricity and transportation parastatals (De Villiers 1984; De Villiers 1986). Reversing its historic policies, the government began moving toward a policy of privatizing its key parastatals.

The transformation of the institutional structure of South African parastatals came about essentially at the very time that South Africa was poised to realize the transition to democracy. Hence, the South African government s privatization proposals were inherently politicized. Whether or not the government intended it, the effect of such privatization would have been to take the parastatals out of the hands of a newly empowered black majority in a democratic dispensation (Horwitz 1992). And indeed, in e arly 1990, the newly legalized African National Congress announced that, though it was no longer wedded to nationalization as a matter of general policy, public corporations and state business enterprises that were privatized prior to political accommodat ion would be prime candidates for renationalization. This political line was voiced shortly after the ANC's legalization and was adopted formally at an April 1990 conference held in Harare, Zimbabwe (Battersby 1990, 3). The ANC's general pronouncement and the specific opposition of the Postal and Telecommunications Workers Association (POTWA, the main union representing SAPT black employees) to the privatization of the SAPT took the wind out of the sails of the privatization gambit.

As a result, "commercialization," which had previously been viewed merely as a preparatory stage to privatization, then assumed importance as a privatization substitute. The government began to view commercialization as a step toward ridding it self of what it had become convinced was a complacent, bureaucratic, top heavy parastatal and as a means of pushing the Post Office toward a market-oriented corporate culture. In place of the command structures of state-owned enterprises, market incentive s supposedly would induce SAPT not only to operate on a proper business footing but would inculcate, it was hoped, a customer-oriented organizational culture in what was widely perceived as a slothful civil service bureaucracy.

3.2 Telecommunications Reform: Telkom SA

South Africa s new telecommunications entity, Telkom SA, became a company formally registered in October 1991 under the South African Companies Act, with the state as the sole shareholder. As a commercialized entity, Telkom could generate profits and pay taxes, received no state subsidies, and was responsible for obtaining its own financing (although this remained subject to ministerial oversight inasmuch as the state was Telkom s sole shareholder). Telkom s management made much of its new status and the change in corporate culture that resulted. It formed a bona fide twelve-person board of directors, consisting largely of high-ranking businessmen (including two nonwhites) and complemented by a coupleof academics. The company decided to operate regionall y (reserving some powers of central management) and appointed individual client managers for important corporate clients. The old SAPT relinquished its joint role as player and regulator, and a very small Department of Posts and Telecommunications acted a s interim regulator. About one hundred people remained in the department while 67,000 went to Telkom.

South Africa s telecommunications sector thus entered an era of liberalization and deregulation. In the early 1990s, however, the actual complexion of this new era was still quite vague. South Africa faced the prospect of fundamentally restructuring its telecommunications institutions, which required the resolution of difficult questions regarding the boundary between monopoly and competitive services, the adjustment of tariff structures, and the coordination of a now lively telecommunications sector tha t previously had very few players. The difficulties of telecommunications restructuring have affected nearly all telecommunications institutions worldwide, but they are particularly pronounced in the advanced industrialized countries (see Duch 1991). What intensifies the complexity and difficulty of the process in South Africa, of course, is that these policy decisions must be considered through the historical lens of apartheid and hence are politicized to an extraordinary degree.

Lest misunderstanding arise regarding this issue of "politicization," I submit that policies on infrastructures are never the technocratic matters they are often portrayed to be. Because infrastructures invariably involve the state, tax monies, and allocative policies that have crucial public goods consequences, policies surrounding them are inherently political. In South Africa, this fact is simply intensified. The historical operation of South Africa s telecommunications, like all state servi ces, was inscribed within the apartheid system, and the inequitable distribution of infrastructure and access to service reflected this fact. The disparities in South Africa s telephone penetration by race are striking, and the politics of the system of s eparate development have meant that there is inadequate infrastructure in the black townships and virtually none in the country s African rural areas. For example, 1989 figures show telephone penetration per one hundred blacks at 2.4 compared with 25 for whites. 13 Number of telephones per 100 residences

___________________________________________________________________

Whites Asians Coloreds Blacks ____________________

Metropolitan Rural

1978 71.5 36.1 19.3 3.3 1.8
1982 83.3 61.5 46.2 24.0 8.3
1987 83.9 72.2 53.2 38.0 13.7
Source: Summarized Report on the Study by Dr. WJ De Villiers concerning the Strategy, Policy, Control Structure and Organisation of Posts and Telecommunications, 1989 [Concise version], p. 3). Note that in this graph, because density is measured by house hold, there is a built-in skew, inasmuch as household size varies by race -- Blacks have largest, followed by Colored, Indian, and White in that order.

This disparity is not simply a matter of skewed income levels. The infrastructure is not fully in place in many areas of black settlement, and there is evidence of a fair degree of suppressed demand among blacks. 14 (1992, p. 9).

Thus, whereas in the industrialized nations the PTTs or regulated telephone monopolies typically marshaled cost-averaging and value-of-service pricing mechanisms to extend service universally, the SAPT historically utilized cross-subsidies to extend serv ice mainly to whites and business. 15 How South Africa deals with this legacy, while accommodating new business needs in the postapartheid era, was the key question facing the sector during the period of transition to majority rule in the mid 1990 s.

 

[rpm]

4.0 THE POLITICS OF TRANSITION

The stated objective of the Post Office Amendment Act of 1991 was to free South Africa s postal services and telecommunications from the perceived burdens of ministerial control. Remaking major socioeconomic institutions, however, is a difficult business . The commercialization of SAPT was a complex political process that required a fair amount of effort to secure appropriate legislation. Several full-scale parliamentary discussions were devoted to the change. 16 It is unrealistic to expect that a technically complex sector like telecommunications can be restructured in specific ways by a general body such as South Africa s Parliament. Hence, it was not surprising that in the early 1990s the determination of many of South Africa s telecommunications issues was deferred.

Two items of enormous consequence were left undecided by the legislation: namely, how the now transformed sector would be structured and how it would be regulated. The central questions confronting any transformation of the PTT system--that is, which are as of service would remain monopolized, which would be competitive and to what degree, and how to adjust tariff structures--were also left unaddressed by the Post Office Amendment Act. Less understandable perhaps was that South Africa s Parliament also di d not determine who would eventually decide these fundamental questions. South Africa s now grossly pared down Department of Posts & Telecommunications was understood to be acting in the role of interim regulator, without, however, having its legal st anding or the scope of its authority spelled out. The 1991 Act to Amend the Post Office Act of 1958 was itself vague and contradictory with regard to the structure of South Africa s post-SAPT telecommunications. On the one hand, section 34 replaced the wo rds "Postmaster-General" with "telecommunications company" and stated that the successor telecommunications company (Telkom):

shall have the exclusive privilege of constructing, maintaining or using, or of authorizing any person to construct, maintain or use, any telecommunications line for the sending, conveying, transmitting or receiving of sounds, images, signs, signals, com munications or other information, and of transmitting telegrams over any such telecommunications line within the Republic or the territorial waters thereof. (Republic of South Africa 1991c, 36).
On the other hand, section 43 stated that:
Notwithstanding anything to the contrary contained in this Act, the Minister may, after consultation with the successor company concerned, and if it is in the public interest, by notice in the Gazette, also authorize any other person to exercise any powe r corresponding with any part of the exclusive power to conduct the postal service or the telecommunications service which has in terms of this Act been transferred to the postal company or the telecommunications company, respectively, on such conditions as the Minister may deem fit (Republic of South Africa 1991c, 44).
Thus, while Telkom seemed to have an exclusive operating privilege and the sole authority to determine the network's use, there was room in the statute for the minister to undermine that privilege. Yet the statute did not clearly give the minister the po wer to compel interconnection with the Telkom network in the event another carrier was authorized to provide a telecommunications service. Moreover, the authority of the Postmaster General vis-a-vis Telkom was at best unclear and at worst minimal. In shor t, the Post Office Amendment Act left a number of central questions unaddressed and an effective vacuum in direction. 17
SinceTelkom is a commercialized company whose sole shareholder, however, is the state, Telkom has in effect answered to two ministers. The "policy" portfolio, held by Minister of Transport and Communications, has been separated from the "s hareholder" portfolio held by the Minister of Mineral and Energy Affairs and Public Enterprises. Because "policy" and "shareholding" are not mutually exclusive categories, in the early 1990s it was not in fact clear who set policy for the sector. 18

The South African government tolerated this lack of clarity perhaps because legislative finality was unlikely and because the interim period was expected to be brief. On the legislative and regulatory fronts, the government asked Pierre Pretorius, a Tran snet advocate (a lawyer admitted to argue before the Supreme Court) who had played a leading role in the commercialization of that parastatal, to draft the legislation on the telecommunications sector. Pretorius's first order of business was to propose a regulatory structure for the sector. South Africa s Minister of Transport and Communications originally favored the appointment of a single regulator in the manner of the British regulatory body Oftel but was dissuaded from taking this course, according t o Pretorius, because South Africa was "too divided a society" and because a group of commissioners would provide greater representation. 19 The working model of the new regulatory body at that time was the U.S. Federal Communications Commission, which oversees American broadcasting, the U.S. spectrum, and telecommunications as a whole.

On the basic policy front, the Department of Posts and Telecommunications contracted with Coopers & Lybrand, Deloitte, the U.K.-based international accounting firm, to conduct an independent study of the South African telecommunications sector. The C oopers & Lybrand study was designed to provide an independent, expert analysis of South African telecommunications to guide the writing of the new legislation. Coopers & Lybrand made available an interim draft of the report, but the government's r eaction was slowed by the then pressing need to attend to the multiparty constitutional negotiations (known as Codesa--the Convention for a Democratic South Africa) in May 1992. In the meantime, the government (probably through the Department of Posts and Telecommunications Steering Group, which oversaw the review) asked Coopers & Lybrand to address several questions regarding the interim draft. The final draft was released publicly in September 1992.

 

[rpm]

4.1 Telecommunications Politics in Transition, Part 1: Assessing the Sector

In 1992, the exigencies of the politics of transition to democracy complicated South Africa s telecommunications picture. At the same time that the government was moving on its own to draft legislation on the structure of the telecommunications sector an d its regulation, it was obliged to discuss at least the rudimentary outline of its proposals at Codesa. This reflected the fact that though the official, formal (National Party) government continued to function, Codesa had begun to take on the trappings of an interim government.

Notwithstanding grumbling from National Party stalwarts about the sanctity of the state and "duly constituted government," consequential state policies simply could not continue to be determined solely by the white government under the fundamen tally new circumstance of the transition to a democratic dispensation. All important policies effectively had to begin going to Codesa. There, however, the key discussions on the subject of a communications regulatory agency centered, perhaps not surprisi ngly, on broadcasting. Because of the inordinate propagandistic power of the National Party stemming from its control of the South African Broadcast Corporation (SABC) (another state monopoly), the ANC-SACP-COSATU alliance focused its attention on breakin g up the SABC and establishing a regulatory authority to ensure fair access to broadcasting for all political parties. The focus on broadcasting made perfect sense given the likelihood of hotly contested democratic elections. In the Codesa negotiations, t he tele communications component of communications regulation was definitely of secondary importance. In terms of its importance to the South African economy, however, telecommunications dwarfs broadcasting. 20

And there is some evidence that the members of the progressive alliance in the Codesa Working Group had little understanding of the technical and economic issues involved in telecommunications or how broadcasting and telecommunications interact. This poi nt was made by Michael Markowitz of Film and Allied Workers Organization, which had been commissioned to comment on the government's communications regulation proposals for the ANC. Markowitz made this claim at a communications workshop sponsored by POTWA in Johannesburg, July 10, 1992. He voiced the opinion that, with regard to telecommunications, the breakdown in Codesa talks may have been a blessing in disguise. In his estimation, the progressive alliance ran a serious risk of agreeing to things it did not understand fully and might well regret later. As for the actual content of the government's first proposals on the regulatory body at Codesa, Markowitz claims that 60-70 percent dealt with the problem of pensions (POTWA Communications Workshop, 1992) .

As if to underscore this fact, the business community voiced concern that telecommunications issues would be given short shrift by any future regulatory body (Financial Mail, June 12, 1992, 83). That same anxiety, albeit perhaps animated by different aim s, was shared by POTWA (POTWA Communications Workshop 1992). Notwithstanding these concerns, the Codesa negotiators had essentially reached an agreement on the terms for a combined telecommunications-broadcast-spectrum regulatory authority. But when Codes a broke down in July 1992, the regulatory authority agreements came undone (Pretorius 1993). 21

The state of South African telecommunications policy in 1993 was characterized by conflict and probably stalemate deriving from the fact that, following the collapse of Codesa, the government went ahead and began to formulate policy without first establi shing a legal framework for the sector and its regulation. Moreover, the government formulated telecommunications policy without engaging an effective consultative process with the ANC and the unions. This in itself was somewhat surprising. In housing, el ectricity, science and technology, and the general economy, national forums in South Africa have been fashioned to serve as negotiating arenas for all relevant parties to hammer out consensual policy recommendations. These forums were pushed by extraparli amentary groups, but the government felt compelled to participate because any long-term policy it might undertake risked being vetoed without the agreement of the forums. The National Electricity Forum, for example, proposed in late 1992 and constituted i n early 1993, found the government participating alongside the ANC, the National Union of Miners (NUM), the National Union of Metalworkers South Africa (NUMSA), the South African National Civic Organization (SANCO), the Association of Municipal Electricit y Undertakings (AMEU), the United Municipal Executive (UME), the Chamber of Business (SACOB), and Eskom to formulate policies to restructure South Africa s electricity sector.

Even in the seemingly more contentious area of broadcasting, the ANC alliance and the government were able to work together and hammer out negotiated agreements. This resulted in the creation of an Independent Broadcast Authority in August 1993 (see Inde pendent Broadcasting Authority Bill, final draft, 1993). However, a parallel process did not gel in the telecommunications sector. Indeed, the absence of such a process in telecommunications led constitutional negotiators to sever the previous connection between broadcasting and telecommunications and negotiate the regulatory authority for broadcasting separately (Pretorius 1993).

In early 1992, BMI TechKnowledge, a market research firm, attempted to put together a telecommunications policy forum, but participation by Telkom and the ANC was diffident and spotty, and the government took no direct part at all. A large convening conf erence was scheduled for October 1992, but in the aftermath of the collapse of Codesa, the ANC alliance refused to participate. BMI abandoned its efforts in early 1993 (Smit 1993). 22 In the meantime, the government received the Coopers and Lybrand Report and asked for public comments within just one month. The ANC issued a blistering attack on the Report, branding it as flawed and denouncing what it perceived as the government's deter mination to restructure South African telecommunications without any broad consultation (ANC 1992). Moreover, the Report's general observation that retrenchments would probably be necessary in Telkom greatly angered COSATU, which had not been consulted by the Coopers & Lybrand consultants (Fanaroff 1993b; Naidoo 1993).

At that point, all the parties began playing out a political strategy. Efforts to get representatives of the ANC and the unions and the government together came to nothing. The South African government then moved ahead with its own agenda, using particul ar aspects of the Coopers & Lybrand Report as its guide.

 

[rpm]

4.1 .2 The Coopers & Lybrand Report

The Coopers & Lybrand Report was a comprehensive study, addressing issues pertaining to regulatory authority and the structure of the telecommunications sector. Beginning with the fact of Telkom's commercialized status, the Coopers & Lybrand Repo rt specifically addressed the problem of institutional roles. Thus, it suggested that Telkom have no regulatory functions whatsoever and that the government's role as sector policy maker be separated from its role as Telkom shareholder. Responsibility for sector policy making would remain with the Minister for Posts and Telecommunications, while the role of government-as-shareholder would be exercised by the Ministry of Finance.

In the interests of ensuring adequate representation and promoting informed public debate, the report suggested that consideration be given to the establishment of a statutory consultative committee for the telecommunications sector made up of representa tives of either consumers or all those with an interest in the sector. The policy-making ministry and the regulatory agency could be required to consult the committee before any major decisions were taken. A proposed Telecommunications Regulatory Agency w ould itself promulgate and administer specific regulations within the context of overall telecommunications policy as set out by ministers. Referring to international trends in telecommunications, the Report recommended that the regulatory agency be set u p as a body independent of the ministry that sets policy (Coopers & Lybrand 1992, 47-70).

In the areas of competition, deregulation, and privatization, the Coopers & Lybrand Report reviewed international trends and supported the principles of competition and privatization in general and in the abstract but recognized that certain service monopolies were necessary, at least for several years. Specifically, the Report recommended that Telkom retain the exclusive right to supply long-distance and international voice telephony for a period of at least five years. The Coopers & Lybrand con sultants recognized that opening South Africa s long distance service to competition would result in widespread "creamskimming" and thus necessitate Telkom's increasing charges for local service by a factor of between two and three. Quite proper ly, Coopers & Lybrand understood that this would be politically unacceptable.

In a sense, a quid pro quo for retention of the long-distance monopoly, the Report recommended that the objective of increased network penetration be formalized in terms of a regulatory obligation on Telkom to meet specified targets for service expansion and provisions made for the financing of such obligations. The Report further recommended that a system of price caps be established for those services in which Telkom retains a monopoly and that explicit regulatory controls on Telkom's prices be institu ted. At the same time, the Report advised that the provision of local service by independent network operators (such as Transtel or Eskom) should not be ruled out. Such provision, however, would not be in competition with Telkom; it should instead be a supplement to the main network (Coopers & Lybrand 1992, 20-26, 81-9 5).

 

[rpm]

In the area of enhanced services, the Coopers & Lybrand Report suggested that the current draft VANs license be issued immediately, subject to negotiations with the industry for further liberalization. However, because of the danger of straight resal e of network capacity, the Report recommended that the resale of voice services should not be permitted for three to five years. Furthermore, the self-provision of short-haul circuits should be allowed subject to defined limits (the Report suggested a lim it of 500 meters). Very small aperture satellites (VSATs) should also be liberalized but made subject to a ban on interconnection with the public switched network.

The Report recommended that provision of the first telephone instrument should be liberalized and the local content requirements relaxed over three to five years. Technical standards would be supervised by an independent body as part of a move toward mor e open, international standards. Finally, in what would soon become the most consequential element of the report, Coopers & Lybrand recommended that consideration be given to licensing two competitive digital cellular networks, among which network sha ring should be permitted to avoid unnecessary duplication of resources. International experience, the consultants argued, shows that competition works to increase penetration and lower market prices in the cellular field. Telkom should be allowed to parti cipate in one of these network-operating franchises, the Report suggested, either on its own or as part of a consortium--but only through a fully separate subsidiary. Air time resellers, the report added, should be introduced to act as intermediaries betw een the networks and the end users (Coopers & Lybrand 1992, 71-80).

Thus, while the Coopers & Lybrand Report recommended some competition to sharpen the incentives for improved performance and held out the goal of privatization as a means to introduce private sector capital into the industry, a close reading reveals that in the short run the Report suggested protecting Telkom as the primary supplier of South Africa s telecommunications services and using a mixed set of regulatory controls to push the company to behave in an efficient, public-interested manner. The on ly recommendation clearly contrary to this general advice was the licensing of two competitive cellular operators. Even here, the Report hedged a bit, advising a prohibition on the provision of connections from mobile to fixed terminal equipment (such as PBXs) in order to protect Telkom. However, the Report clearly did not adequately analyze the consequences to Telkom and the public switched telephone network of its cellular recommendations (Coopers & Lybrand 1992, 76-78). Sensitive perhaps to the ver y real politico-economic tension between the need to expand South Africa s basic network in the postapartheid period and the need to liberalize VAN services, the Coopers & Lybrand Report plotted three potential scenarios that described alternative dev elopment paths for South Africa s telecommunications sector:

Scenario 1, the "base case," essentially followed Telkom's five-year plan for modest service expansion and investment, extended forward to 2002.

Scenario 2, labeled "network expansion," proposed to increase the availability of South Africa s basic telephone service in an aggressive manner, increasing access lines to 19.6 per one hundred urban inhabitants versus 14.1 in the base case (sc enario 1), and 6.2 access lines per one hundred rural inhabitants versus 3.5 in the base case. Although there would be an increase in both local and national call charges, access would be cross-subsidized mainly by business through higher rental and conne ction charges. International charges would decrease, however.

In scenario 2, Telkom would not face competition in voice telephony, but its profitability would be the lowest of the three options, primarily due to the significantly higher levels of capital investment. Some financial restructuring would be necessary t o carry out the aims of scenario 2. Indeed, in order for the network expansion scenario to work, the Report suggested the necessity of a capital injection of R4 billion in 1993-94 to retire an equivalent amount of long-term debt. This cash could come from the government or by privatizing Telkom.

Scenario 3, the "competition" option, rebalanced all prices over five to seven years to reflect actual costs, after which competition in voice telephony is introduced and prices presumably decline thereafter. Under this scenario, business price s decline furthest and fastest. To increase the availability of basic service, public pay phone penetration would be increased substantially to nearly 300,000 by the year 2002 versus 82,600 in Telkom's base case (scenario 1).

In scenario 3, access line penetration plots out to 13.5 per one hundred urban inhabitants by 2002 and 3.5 per one hundred rural inhabitants--quite a bit lower than the "network expansion" scenario (scenario 2). Competition in all services is a ssumed to be a spur to productivity gains. Telkom would be in the best financial shape under this scenario.

To the extent that the Coopers & Lybrand Report offered recommendations among the scenarios, it suggested that both the "network expansion" (scenario 2) and "competition" scenarios (scenario 3) were better than Telkom's base case plan (scenario 1). Telkom's plan was judged to lead to only a limited expansion of levels of access and even that depended on network utilization increases that approached operationally infeasible levels. It seems reasonable to conclude that the Coopers & amp; Lybrand consultants addressed what they understood to be the two main politico-economic policy positions with regard to telecommunications--and in their Report framed these positions within the discourse of expertise.

In this sense, the Coopers & Lybrand Report does in some ways represent a kind of generic, framed debate between the ANC alliance and the National Party over the future of South African telecommunications. And this perhaps was the Report's real impor tance. It presented a version of the basic progressive demand for network expansion and framed it within acceptable boundaries. Given that the ANC alliance had neither the wherewithal nor the capacity to generate its own study, the Coopers & Lybrand R eport essentially became a baseline from which to assess South Africa s telecommunications policies and the debates about them. Thus, any demand that Telkom "serve the people" had to confront the specific realities of Coopers & Lybrand's sce nario 2.

At the same time, any call for a competitive telecommunications system for South Africa must address the demands for basic service by, at the very least, vastly increasing the number of pay phones and keeping call charges within reach of the poor. In oth er words, the Coopers & Lybrand scenarios may not be the only options, but they likely assume the status of a "gold standard."

 

[rpm]

4.2 Telecommunications Politics, Part 2: The Cellular Controversy

In certain respects, the Coopers & Lybrand Report reflected a wariness that resolving the conflicts in South Africa s telecommunications sector would be a difficult political process. This wariness, however, was evidently not shared by the South Afri can government. Advocate Pretorius, charged with drafting the enabling legislation, wanted to set the legal structure in place before policies were determined. Short of a legal structure, Pretorius favored a consultative process for the determination of p olicy.

Though seemingly placed in a powerful position, Pretorius was essentially ignored. The government went ahead with a strategy to revamp South Africa s telecommunications sector without a legal structure in place and without a consultative process. The veh icle for the revamping was the decision to establish two cellular network operators and send these licenses out to tender. Cellular thus became the key battleground in the South African telecommunications debate.

The visible driving forces in this movement were Minister of Transport and Communications Welgemoed and Eugene van Rensburg, head of the Policy Unit on Public Enterprises of the Ministry of Mineral and Energy Affairs and Public Enterprises. 23 This route was in marked contrast to that taken in other contested sectors. For instance, in 1993 South Africa s Department of Home Affairs was sitting on eighty to ninety applications for broadcast licenses, but it was decided not to make any awards unti l the broadcast authority was put in place (Pretorius 1993).

South Africa's cabinet met in early 1993 to decide how the proposed cellular mobile telephone system should be organized and regulated. With expressions of investment interest from big-time foreign operators--perhaps representing the largest single forei gn investment in postsanctions South Africa--dangling in front of them, cabinet ministers went ahead and authorized two cellular licenses. The parties pushing for the quick authorization of cellular service pointed to the interest by foreign investors and argued that the timing was opportune and not to be squandered (see Sergeant 1993a; also Knott-Craig 1993b). The licenses were put out to tender in April 1993, and the Minister of Transport and Communications appointed Ters Oosthuizen as the "Regulat or." Oosthuizen, an advocate who for many years worked for Eskom, was to oversee telecommunications in general and the cellular tender process in particular. Oosthuizen subsequently asked the minister to name him Postmaster General because he had no staff or legal standing as Regulator. However, by becoming Postmaster General, Oosthuizen unambiguously identified himself with the government and thus its old structure and nonconsultative methods (Hainebach 1993).

The ANC and COSATU, joined by a group called the Cellular Telephone Consultative Forum, immediately charged that Oosthuizen had been appointed outside of both legislation and the Negotiating Forum (the resurrected Codesa constitutional negotiations). The ANC alliance claimed that the licensing of cellular telephony represented a unilateral restructuring of the industry, a form of privatization through the back door (see Fanaroff 1993a). In fact, some of the government's own pronouncements support that in terpretation. Policy Unit chief Eugene van Rensburg, for example, was quoted as saying that the cellular policy can be taken as clear indication that the government favors competition (Financial Week, January 21-27, 1993).

Another explanation of the government's actions put politics at center stage. According to one report, a senior government source was quoted as saying that the National Party government saw that it had the opportunity to provide telephones to the masses quickly through cellular service, and the Party had no intention of waiting and thereby permitting the ANC to reap the political benefits (Sergeant 1993b). 24 Other observers invoked a political explanation of the more traditional kind, arguing that Welgemoed was a minister "of the old school"--that is, of the imperious apartheid stripe--who would never allow "interlopers" to make policy or interfere with his authority. 25

Whatever the full explanation for the South African government's actions, in 1993 it proceeded to set the terms of application for the cellular licenses and appoint a tender review board to review the applications. A R50,000 tender application fee was as sessed in order to discourage all but the most serious bidders. The basic terms of the fifteen-year license were that each operator had to pay an initial basic fee of R100 million plus an ongoing license fee of 5 percent of its net revenue. Also, R5 milli on in annual radio fees had to be paid to the Postmaster General along with R1 million a year for each 10 MHz channel granted to the operator. Some of these fees were designed to finance the Regulator, but the R100 million basic fee would go straight to t he Exchequer--it was not earmarked, that is, for expanding telephone service to those without access (Financial Mail, April 16, 1993). Among other things, applicants had to specify the extent to which their choice of technology would lead to high volumes and low costs, how they would support South African industry, and how they would provide service to poor communities. 26

Telkom was quickly awarded half of one license. It entered into a consortium with U.K.-based Vodafone and the Rembrandt Group of South Africa to form a separate subsidiary, Vodacom. Three applicants vied for the second license. Mobile Telephone Network, considered by most accounts the favorite, was a consortium of M-Net (the South African pay television provider), U.K.-based Cable and Wireless, NAFTEL (an association of black businessmen), and Transtel. Cellstar Cellular Networks was a venture of Anglova al's Grintek Limited and Telecom Finland. The Reunert Group applied for the license in a venture that combined Barlow Rand with the Deutsch Bundespost Telekom. Both the Reunert and Cellstar ventures left room for black participation, but at the time the t ender applications were submitted there were no takers. 27 Vodacom: Telkom (50 percent), Vodafone (35 percent), Rembrandt (15 percent).

MTN: M-Net (30 percent), Cable & Wireless (30 percent), Naftel (30 percent), Transtel (10 percent).

Cellstar: Grintek [Anglovaal] (51 percent), Telecom Finland (11 percent), Finnfund (11 percent),

Black business [unallocated] (27 percent).

Reunert: Barlow Rand (34 percent), Deutsche Bundespost Telekom (34 percent), Black business [unallocated] (32 percent).

 

[rpm]

In 1993, the African Telecommunications Forum, a loose grouping of black organizations and entrepreneurs, made inquiries about obtaining a cellular license, though it did not submit a genuine tender bid (in part because of ANC opposition to the tender pr ocess). The parties to the African Telecommunications Forum included the ANC-aligned Thebe, Afritel, Suntel, MIT, and NIT, and the Forum enjoyed the backing of the National African Federated Chambers of Commerce (NAFCOC). Afritel in particular forged link s with the U.S. telecommunications giant, AT&T, although the U.S. investors would not commit until all state and local authority sanctions were dropped (Financial Week, September 23-29, 1993; Young 1993).

In late September 1993, the South African government announced Mobile Telephone Network as the winner of the second cellular license. Vodacom's system would be the Global System for Mobile Communications (GSM) standard, and Mobile Telephone Network propo sed GSM as well. Quite apart from the licensing controversy, a minor storm arose over the technical standard. The part of the spectrum allocated to South African cellular was the 900 MHz, eliminating the American Mobile Phone Standard system, which operat ed in the 800 MHz spectrum. The fact that the tender documents specify a R20,000 a year fee for each 200 KHz channel led some to question whether the South African government had already chosen a standard--because only the GSM operates in 200 KHz blocks ( Financial Mail, April 16, 1993).

Several observers, including Postmaster General Oosthuizen, indicated that well prior to the tender bidding Telkom had already committed itself to GSM and had installed some base stations (Oosthuizen 1993; also Maepa 1993). Moreover, Telkom's commitment to GSM, in the judgment of the chief executive of Plessey, stemmed not from its technical characteristics per se but once again from Telkom's historically close relations to its main equipment supplier, Siemens--the company that would bring in GSM technol ogy from Europe (Temple 1993). 28 GSM was faulted by some as an expensive cellular system, with high prices for handsets that are not very lightweight and probable call charges several times those of the current phone network. Others, notably Alan Knott-Craig, chief executive of Vodacom, were forthright in their defense of GSM. The key, Knott-Craig argued, was that GSM allowed the separation of service from the actual telephone. Thus, GSM cellular call boxes could go a long way toward serving underprovided communities. Vodacom's Community Tariff Plan subscribers would use a prepaid SIM card, enabling them to make calls from any GSM telephone and access a centralized voice mail system. The voice mail feature would obviate the need for the individual ownership of a telephone. In addition, V odacom planned to franchise the GSM Community Telephones to members of underserved communities, creating the conditions for telephone entrepreneurship (Knott-Craig 1993b). 29 Of course, the complexity to the user of the SIM Card and voice mail led others to doubt the very claims put forward for GSM (Davies 1993). 30 While the ANC and COSATU did not condemn GSM per se, they bitterly opposed the existing cellular tender process. In 1993, ANC Information Systems head Andile Ngcaba indicated that he believed the cellular applicants were simply using the language of unive rsal service to win the tender. The ANC wanted cellular to be a separate, autonomous parastatal service offered by Telkom in a sector rationally planned to take advantage of all existing communications infrastructure. Such a structure would be the best me ans for providing universal service (Ngcaba 1993a; Ngcaba 1993b, 17). 31

The cellular controversy thus became a major source of conflict between the government and the ANC alliance. The government argued that it would not stop the tender process; that applicants had spent millions of Rands in a good faith effort to apply for licenses, and suspending the process would lead to a myriad of law suits. The government also argued that suspending the process would jeopardize foreign investment (Business Day, September 13, 1993; Oosthuizen 1993).. 32 Even as the South African government was going ahead with the tender process and the ANC alliance was denouncing it, efforts were being made to bring the two sides together. A meeting was set up between Postmaster General Oosthuizen and ANC Information S ystems head Ngcaba, but it never materialized. Ngcaba, according to one inside player, indicated that if the ANC alliance could get a few of its people appointed to the Tender Board, it would probably go along with the cellular licensing process. That, to o, however, did not happen.

The ANC alliance then called for a moratorium on the tender process. In September 1993, the politics of the cellular tenders became quite heated, with COSATU threatening strikes and the ANC threatening to revoke the licenses when it came to power. ANC Pr esident Nelson Mandela met with President de Klerk over the cellular license controversy, and ANC Secretary-General Cyril Ramaphosa led a delegation to meet with Public Enterprises Minister Dawie de Villiers (Makhanya 1993a, 18). In September 1993, Ramaph osa was quoted as saying that a future government would immediately review, and perhaps cancel, the cellular licenses if the government went ahead and issued them (Citizen, September 15, 1993).

Interestingly, the cellular controversy cooled off because another telecommunications brouhaha intervened. Moving ahead in other areas, the government tabled a new bill before Parliament to amend the Post and Telecommunications Act. While much of this bi ll simply clarified the power of the postmaster general, the ANC alliance again vehemently objected to the government s lack of consultation in what appeared to be a bid to restructure South African telecommunications. And in the context of the struggle o ver the cellular licenses, the ANC interpreted the bill s clarification of the postmaster general s authority as a cynical maneuver to deregulate Telkom (Ngcaba, 1993b, 17). In exchange for the government s agreement to hold off on the new Post and Teleco mmunications amendment bill, the ANC agreed to back down on its opposition to the granting of the cellular licenses (Business Day, September 30, 1993).

 

[rpm]

5.0 THE STATE OF THE NETWORK DURING TRANSITION

The government to be popularly elected in 1994 stood to inherit a relatively large telecommunications network consisting of 3.7 million access lines--the twenty-fifth largest network in the world (Department of Posts & Telecommunications 1992, 49). T his total had increased from 3.315 million in 1991. South Africa has 38 percent of the telephones in Africa and the highest density of telephones on the continent (Siemens 1992, 16; Telkom 1991, 1).

South Africa's network is also relatively sophisticated. It has a high degree of digitalization and consists of an increasing percentage of optical fiber. As we have seen, SAPT began the process of digitalization early, in 1978-79, in part because it ran into trouble with its electromechanical switches just at the point when large foreign manufacturers were introducing digital technology and in part because digital was judged by SAPT management to be a better technology for meeting the huge existing dema nd. 33

This point raises the question as to whether the conclusions of the De Villiers Report were warranted. The SAPT was the last of de Villiers' investigations. According to WJ Taylor, former Deputy Postmaster General for telecommunications, after finding re al problems with the electricity and transportation parastatals, de Villiers came to the Post Office investigation with preconceived notions and, by then, a set agenda. Grossly ignorant of telecommunications, de Villiers insisted that SAPT's decision to g o digital had been a mistake and was a gross instance of goldplating. Yet nearly universal expert opinion argues that digital is more flexible and in the long run cheaper than electro-mechanical systems (Taylor, 1992). Siemens, Temsa, and Altech were awar ded long-term contracts for the supply of this equipment and began installing digital exchanges in 1980.

As of March 1993, there were 1,586 exchange sites in South Africa s automatic telephone network of which 929 had digital exchange equipment--a digital penetration rate of 58.5 percent. Of the over 3.5 million access lines in operation in 1993, about 56 p ercent terminated on digital equipment. Moreover, this rate was planned to increase to 64 percent by 1995 and to more than 80 percent by the year 2000. The penetration of digital transmission systems is even higher, with 100 percent in South Africa s larg e metropolitan areas and 78 percent in the rural areas. In the early 1990s, the network was running at about 80 percent of capacity (Lachenicht 1991; Telkom 1991; Telkom 1991; Telkom 1993).

Prior to 1969, South Africa s international telephone connections were carried by some twenty high frequency circuits and two telegraph cables. In 1969, the SAT-1 submarine coaxial cable, running from Cape Town to Portugal, began operation and in Decembe r 1975 was supplemented by a satellite earth station. A new undersea fiber-optic cable (SAT-2), due to be commissioned in 1994, would carry a 565 Megabits per second transmission--providing the equivalent of 7,680 voice channels. SAT-2 would link South Af rica to international telecommunications nodes in Madeira and the Canary Isles (see Proceedings 1991). In the mid-1990s, a modern telex service with over 30,000 connections linked all of South Africa s major population centers--although, as in all nations , facsimile or fax service was displacing telex, and telex investments consequently became a major loss. South Africa was one of the first countries to introduce a public videotex service (Beltel), which by 1993 included an x-400 protocol electronic messa ge handling service. By the end of March 1992, the number of registered Beltel users was nearly 30,000, but the service has historically operated at a considerable loss (BMI TechKnowledge 1992, 99). In 1983, an international packet-switching network (Sapo net-P) replaced an analog network and by 1993 linked South Africa to thirty-five countries. In 1986, a digital point-to-point service (Diginet) for companies needing to transfer large volumes of data at high speeds was introduced.

 

[rpm]

In 1993, leased lines, including high-capacity 2 Megabits lines, became available in South Africa but could only be purchased from Telkom. Under 1993 rules, a link between two offices of the same company on either side of a public road were to be provide d by Telkom. Neither third-party traffic nor bypass was permitted. South Africa s cellular phone service, based on a variant of the German C450 analog system, had been operating since 1986 but by 1993 had only about 13,000 subscribers because of extremely high handset prices and the lack of available frequencies .

South Africa s liberalized private Mobile Radio and radiopaging services had also grown considerably by the early 1990s, but they too were hampered by a dearth of available radio frequencies (largely because some 62 percent of the frequencies allocated b y the International Telecommunications Union for Land Mobile Radio use in the early 1990s were occupied by the South African Defence Forces, although the sharp reduction in the military over the next years would create opportunities for reallocation of fr equenices) (Department of Posts & Telecommunications 1992, 14; Coopers & Lybrand 1992, 76).

By the mid-1990s, various customer premises equipment, including PBXs, modems, fax machines, and cordless telephones, had been deregulated in South Africa for several years--subject to type approval by Telkom. In addition, Telkom provided its own small b usiness telephone systems and data modems in competition with privately supplied PABX systems and data modems. By 1993, internal wiring within buildings had been largely liberalized for several years as well. The liberalization of customer premises equipm ent had limits, however. Telkom retained exclusive rights to provision of the first telephone on customer premises, and the approval of attachments involved local content requirements until 1996.

In general, the quality of South Africa s telecommunications service is good and service is relatively cheap. For example, Telkom's call completion rates (just over 95 percent) and speed of fault clearance (77 percent cleared within one working) are near the level of other industrialized countries. In white and urban areas, the average waiting time for connection in 1990 was thirty days (down from sixty-seven days in 1989), although in black areas waiting time for connection is much higher and extremely variable (Coopers & Lybrand 1992, 14). Notwithstanding these statistical measures, in the early 1990s large users reported unhappiness with Telkom s response on fault clearance and connection times and frustration with Telkom bureaucracy generally (Da vies 1992; Paul 1992). Moreover, the service waiting list has been growing: between 1990 and 1991, the waiting list had grown 14.56 percent to an official total of 125,448 and by March 1994 to an estimated 134,000 (Department of Posts & Telecommunicat ions 1992, 12; Plessey 1995). 34

The waiting list reflects actual applications, hence it mainly reflects demand in areas where there is some infrastructure. The list does not adequately take underdeveloped areas into account.

By the early 1990s, Telkom also offered telephone information services, but the problems that plagued it in the early months of 1992 were revealing. Telkom's response to widespread customer complaints about grossly inaccurate bills for example, was first to deny that there was any problem at all and then to insist that complainants pay their accounts (some reading in the thousands of Rands) while Telkom investigated--or risk having telephone service cut off. Finally, amidst great hue and cry, Telkom's ma naging director conceded that computer error and/or service theft were responsible for incorrect billing. The "087" debacle (from the number for information services) lent credence to the view that, for all the talk about a new customer orientat ion, Telkom's corporate culture had not yet changed. The entire episode smacked of the arrogance and unconcern of a single supplier and left Telkom with bad debts of some R77 million and contributed to write-offs of R136.9 million (Telkom 1993, 3), a figu re that would grow to R207 million by 1994. 35

The press minced no words in attacking Telkom. Typical was this editorial from the Cape Times. "The telephone service was far from perfect when run by the Post Office, earning its fair share of complaints from the public. But never did it arouse so enraged and sustained an outcry as Telkom has within a short time of taking over the service....If this is the result of what is laughably called privatisation or commercialisation, then the telephone service should be re-nationalised without delay. Any p rivate company which abused its clients in this way would quickly go out of business. But Telkom has no competition. Whatever similarity exists between its monopolistic stranglehold on telephonic communication and private enterprise is in the imagination only of the government which foisted it on the public" (Editorial, 1992, p. 8).

International price comparisons of a basket of telephone services with industrialized countries (based on 1989 tariffs and exchange rates) show that Telkom's tariffs for both business and residential customers as well as residential line rental and conne ction charges were low by international standards. In addition, unlike many countries, South Africa s business customers not charged higher prices than residential customers, and business line rental and connection charges are thus even lower by internati onal standards than those for residential customers. On the other hand, its international long-distance tariffs were comparatively high. For instance, charges for calls to the United States were some 30 to 40 percent greater than U.S.-originated calls to South Africa (Financial Mail, June 25, 1993). National long-distance charges were in line with those in other industrialized countries. Call charges account for approximately 50 percent of Telkom's total revenue (Coopers & Lybrand 1992, 11-17). Hence, South Africa s cross-subsidies as of 1994 can be understood as follows: international--and to a lesser degree national--long-distance charges subsidized local service (although this is not direct business subsidization of residential while calls greatly subsidized rentals. According to the De Villiers Report , the actual cost of telephone rental per month in South Africa in 1987-88 was about R28, though the actual rental charge was R15. Thus, telephone rentals have been heavily subsidized by higher call tariffs. However, according to conventional wisdom (larg ely based on the De Villiers Report ) a large proportion of subscribers (particularly new, usually nonwhite subscribers) make very few calls per day. As a result, telephone revenues rested on a narrow base of subscribers. According to the De Villiers Report , 6 percent of subscribers (of whom about 78 percent are business and 22 percent residential telephone subscribers) contributed 50 percent of South Africa s total telephone revenue in the late 1980s (De Villiers 1989, 25, 8). 36

 

[rpm]

6.0 TELKOM'S CURRENT CONDITION AND FINANCIAL HEALTH

In 1994, Telkom had a monopoly in South Africa s various service areas, the most significant of which was the running of the public switched telephone network. Telephone instruments, telex/teletex terminals, and data network terminating units connecting into the public switched network remained Telkom's exclusive business in 1994. Private lines that were not confined to the boundaries of a client's premises were still monopolized by Telkom. With a payroll of some 62,000 people and assets valued at nearly R15 billion, in 1994, Telkom had an annual operating expenditure of more than R5 billion. 37

Notwithstanding Telkom's commercialization and its trumpeting of a new corporate culture, several factors clouded its prospects. Over the period 1987-91, Telkom's operating revenues grew by an average of 1.9 percent per year in real terms, whereas operat ing expenses grew nearly twice as fast, by 3.7 percent. As a result, Telkom's gross profit margins declined significantly, from 42 percent of turnover in 1987 to 30 percent in 1991(Coopers & Lybrand 1992, 16-18), a figure which would remain unchanged through 1994. Moreover, despite expected moderate surpluses in telephone services (R790 million) for the year ending March 1992, very large losses on telex (R230 million), gentex (R94 million) , and postal services (R268 million) largely wiped out the surplus (Department of Posts & Telecommunications 1992b, 11).

Telkom's debt to equity ratio 1.8 made it unlikely that it could easily raise more debt financing without state guarantees, although Telkom had substantially improved its financial position since the early 1990 s. 38 Indeed, Telkom's capital expenditure since its commercialization (2.2billion in 1994) has come from internally generated funds only (Telkom 1993, 4). At the same time, an outflow of funds from the Post Office Savings Bank--a net amount of R521 million fro m 1989-90 to 1990-91--represented a loss of access to cheap capital, 64 percent of which had historically been invested in telecommunication assets (Department of Posts & Telecommunications 1992a, 6). In 1991-92, Telkom's total financing costs were ex pected to be R1.7 billion compared with a net operating profit of R2.4 billion (Coopers & Lybrand 1992, 16-18). Approximately 20 percent of Telkom's budget was allocated to service its borrowings; interest payment coverage stood at only 1.9x (Fleming Martin, 1995, p 8). According to Telkom s 1994 report, its official net interest bearing debt stood at R8.8 billion. The terms of its commercialization entailed government-imposed constraints on Telkom with regard to the retrenchment of staff (namely, no retrenchment as a result of commercialization), real increases in tariffs (which remain subject to the approval of the Department of Posts and Telecommunications in its role as interim regulator), recourse to external finance, and the honoring of long-ter m contracts with local suppliers. In addition, in 1990-91, Telkom was obliged to contribute R367 million to the postal deficit, with the expectation that it would be responsible for R500 million annually until 1995-96, to be met through taxes and dividend s paid to the government. As if this were not enough, the Telkom Pension Fund was found to be short R1.59 billion and would have to be fully funded, though by the end of 1994, Telkom had succeeded in reducing the deficit to R1.1 billion.

Thus, even as revenues were declining in the period between 1 October 1991 (when it became a company) and March 1995 Telkom found itself contributing R209 million to the postal deficit, R446 million to the pension fund deficit, and R410 million in taxes (Telkom 1993, 56). Telkom's poor financial status translated into a significantly lower amount of capital investment--precisely at the moment when the end of apartheid dictated a vast expansion of the network. In the period 1975-86, telecommunications inv estments as a percentage of revenue typically averaged around 45 percent, but beginning in 1987 that figure dropped precipitously and in 1994 registered just 27 percent. At 2.227 billion, 1994 nominal values of capital investments in the network were 51 p ercent higher than the 1987 levels of R1.471 billion--but allowing for 15 percent average annual inflation, this was in fact negative real investment growth overall.

Moreover, according to Telkom s 1993 annual report, it spent R1.8 billion on capital programs, representing just under 17 percent of turnover (Telkom 1993). Telecommunications investments as a share of GDP, which peaked at 1.07 percent in 1986, fell to . 64% in 1994 (ITU,Draft African Telecommunications Indicators, 1996). This parlous state was compounded by Telkom s depreciation rates, which, at an average of 4.7 percent between 1980-81 and 1990-91, were too low to continue the modernization of South Afr ica s network--particularly in those areas where it would be subject to competition (Department of Posts & Telecommunications, various years). In data and value-added services, equipment changes quickly. Without provision for replacement, Telkom canno t buy new equipment and easily retain business customers. Telkom recently went to a fifteen-year depreciation schedule for transmission and switching equipment, and ten to twenty years on cable, but some knowledgeable observers viewed even the new depreci ation rate as too low (Telkom 1993; Hainebach 1993b).

In short, in the mid-1990s Telkom faced a difficult future. On the one hand, the end of apartheid imposed new equity-based demands on it to expand the basic telephone network to populations and areas historically denied access. The black Postal and Telec ommunications Workers Association (POTWA), a mainstay of the powerful Congress of South African Trade Unions (COSATU), strongly supported the expansion of the telecommunications infrastructure both for reasons of equity and to safeguard its members' jobs.

 

[rpm]

In response to the demand to expand the network, during the 1992-93 fiscal year Telkom made a special allocation of R40 million for purchasing and commissioning "Community Telephone" facilities in remote and rural and inaccessible areas, such a s squatter camps. Its stated objective to install 250,000 phones a year between 1992-93 and 1997-98 (with emphasis on "community phones") would be accomplished by the installation of a high frequency radio network (Rurtel) and the establishment of base stations throughout the country. Rather than attempting to provide each household with a telephone, Telkom would aim to install phones in all businesses and set up call boxes or card phones within "reasonable access" of everyone (Stracha n 1992). Nevertheless, the R40 million earmarked for community telephony is a paltry sum when one considers the pressing needs of South Africa s disadvantaged and Telkom's R2.14 billion capital budget for 1992. 39 years. The research of Mike Morris and SE Stavrou establishes the workability of authorizing a township telephone customer to operate an agency telephone out of his/her house. Temsa constructed a special meter that indicates the real-time cash charges of any call, thereby making charges transparent to the customer. According to Karl Hartyani of Temsa, Telkom made no commitment to the agency telephone (Morris & Stavrou, 1991; Hartyani, 1993).

On the other hand, South Africa's businesses, feeling themselves restricted under a traditional, natural monopoly telecommunications structure, wanted enhanced and value-added services to be liberalized. Under the Post Office rules, it was illegal, for e xample, for an automobile dealer to use the data network of his or her supplying company to get information since this constituted transferring data outside of SAPT provision or control. The patent economic irrationality of such rules in an age of rapid i nformation transfer as well as the pressure brought by South Africa s National Telematics User Group (NTUG) prompted Telkom to allow companies to register a Value Added Network (VAN) services company in the early 1990s. Yet these agreements forced a VAN p rovider to become a separate company whose network and revenues were audited by Telkom in order for Telkom to ensure that the VAN was in fact a VAN and not a second telecommunications operator in disguise. The difficulties and improprieties posed by Telko m acting as regulator of competitors on the edge of its monopoly prompted an outcry among VAN providers, and in the early 1990s the VAN license recommendations again came up for negotiation (Davies 1993).

Telkom's actions surrounding the VANs issue, if blunt and heavy-handed, did reflect a legitimate concern, however. If and when value-added services are liberalized, Telkom must rebalance its general tariffs in order to compete. In 1993, telephone traffic accounted for over 92 percent of South Africa s telecommunications income, and that traffic was growing at about 7.5 percent per year (although, with the recession of the early 1990s, telephone traffic actually declined 1.5 percent in the twelve months p rior to April 1993 [Telkom 1993, 3]). However, it is in leased circuits and enhanced services that the growth potential for South Africa s telecommunications sector really lies. Leased circuit growth for 1991-92 was estimated to top the previous year by o ver 20 percent (Department of Posts & Telecommunications 1992b, 11; ITU 1992, 353). And it was precisely in these areas that Telkom faced potential competitors. Value-added voice and data services that allow operators to employ lines leased from Telko m for third-party traffic were a contested area in the early 1990s and represented a potential threat to Telkom. Large users want the freedom to carry third-party traffic and supply all services (including internal voice traffic) in an economically centra lized way. Telkom, essentially acknowledging this as inevitable, wanted time to establish fair competition in value-added services and create a "level playing field." This could not mean anything other than moving toward cost-based pricing in al l areas of service (see Knott-Craig & Hanekom 1990; National Telematic User Group 1991). Such a move, however, pointed to a fundamental tension between the need to rebalance tariffs so as to liberalize VAN services for business and the pressing need t o expand the network for basic service.

In addition, because of the immense amount of data traffic they pass every day, South Africa s large users (banks in particular) were very leery of total dependence on the Telkom network and thus keen to have a redundant network (or networks) available. Waiting in the wings were two and perhaps three organizations that already had the capacity to provide alternative networks. Eskom, the electricity supply monopoly, and Transtel, the now for-profit subsidiary of the commercialized transportation monopoly Transnet (formerly South African Railways and Harbours), each have large internal networks that could be made available to third parties on a point-to-point basis. Eskom operates a sophisticated microwave network and powerline carriers. Eskom and Transtel could also act as contractors for building new capacity to order, connecting their clients' private networks. 40 The South African Broadcast Corporation's signal distribution division could also, in theory, evolve into a carrier of point-to-point third-party traffic. Telkom s most likely potential competitor is Transtel, which in 1993 already operated a nationwide h igh-capacity digital microwave network consisting of several 34 Megabit systems, a trunked radio network, and a 140 Megabit fiber-optic network in the PWV and Natal (Transtel, no date). Because South Africa s main roads and town arterials follow rail line s, Transtel's radio network extends to geographic areas not covered by the Telkom network. In addition, Transnet s connection to the PanAmSat would reduce the difficulty of reaching remote rural areas. In capacity, Transtel's network is hardly a rival to Telkom's. However now that it has been transformed from a state enterprise into a commercialized business, Transtel is looking to maximize the utilization of its facilities in profit-making activities. 41

Finally, due to competition from international discount carriers, in the mid-1990s Telkom was under pressure to offer international telephone connections at U.S. carriers' rates . Several international resellers, including Diners Club, IDT, and Word-Phone begun operations in South Africa in the early 1990s, and Telkom responded by cutting its own international rates (Financial Mail, June 25, 1993). The reduction of rates by approx imately 30 percent threatened to erode Telkom's earnings in overseas long distance, further undermining its ability to cross-subsidize the expansion of the basic network.

 

[rpm]

7.0 PRIVATIZATION AND CURRENT POLITICAL TRENDS:

The ANC-led Government of National Unity inherited the issues left unresolved since the commercialization of the Telecommunications industry. First, the direction of the industry had remained uncertain. Telkom was the monopoly provider of mot telecommuni cations services, but a liberalized cellular market had claimed 400,000 subscribers by the end of 1995 -- 10% of all the lines Telkom operated. Potential competitors in the equipment manufacture, cellular and service industries demanded greater openness. And despite heightened demands for "universal access," no consensus emerged about the definition of the term.

Second, Telkom had essentially remained unregulated since its founding. Tariff setting and labor market policies, for example, had remained unchanged since the apartheid era. A new regulatory body was needed to create a framework for the growing competit iveness of the market, determine its limits, if any, and chart the appropriate course for the social and economic goals of the GNU.

Third, the role of Telkom itself required to be clarified. The "commercialization" of Telkom was no more than a partial step taken as a political compromise. Issues such as its future ownership, its financing and its obligations towards its wor k force, consumers and society remained unresolved. With its poor service record and inadequate investment in network expansion, Telkom was an unsuccessful agent of the government s goals. At the same time, its precarious financial position and inefficien t operations rendered it a poor prospect for success in a competitive business environment. The Cooper s and Lybrand report framed the issue as, essentially, a choice between social responsibility and fiscal responsibility; the government could either exp and its network rapidly to provide service to disadvantaged people, at a high cost, or it proceed less aggressively, improve the financial performance of Telkom and slowly establish a viable competitive environment.

Responding to political pressure from business, consumer groups and labor, the ANC would attempt to pursue both objectives. Three events consequent to the 1994 elections began to move telecommunications policy in [this direction. The long-delayed establi shment of a National Telecommunications Forum (NTF) finally engaged government, business, labor, and civic organisations in the consultative processes that stakeholders in other sectors had established years earlier. A broad, Keynesian-inspired developmen t plan for post-apartheid South Africa (the Reconstruction and Development Programme or RDP) set goals for telecommunications development, including a specific goal that telephones should be provided to all existing schools and health clinics within two y ears (Republic of South Africa, 1994). And the new (ANC) Minister for Posts, Telecommunications and Broadcasting, Dr. Z. Pallo Jordan, put in motion a Green Paper/ White Paper process to develop telecommunications policy.] The Green/White Paper process, w hich would take place outside the old regime-identified Department of Posts and Telecommunications, tied the politics of telecommunications reform to the more open, consultative forums and mechanisms of which the NTF was the guiding spirit.

7.1 The Consultative Process

The Green Paper was published in early July, 1995 in several languages and put on the World Wide Web as well (Ministry of Posts, Telecommunications and Broadcasting, 1995a). By October, when the period for submissions finally closed, 131 submissions had been made, amounting to over 4000 pages of commentary. There were many points of consensus, including the need to expand the sector through new sources of investment, on the need for operators, service providers, and equipment manufacturers to commit to r etraining and redeployment of staff, that universal access is a reasonable moment on the way to the goal of universal service, that an independent regulator must supervise the system. On the major questions of market structure and ownership, there was a wide range of opinion, from the maintenance of state contro l to full competition and privatization. While most of the value-added network players, some important users, and potential competitors to Telkom (both South African and international) stressed that full-scale liberalization and the privatization of Telko m had to occur very soon, the labor unions and an important black business group (the Foundation for African Business & Consumer Services) held out strongly for minimal change and continued state-control and ownership of Telkom. Telkom itself proposed a fairly long period of exclusive concession to enable it to mobilize capital and its large workforce to extend the network to the disadvantaged and rebalance tariffs to prepare for competition. It also advocated that it be permitted to take on strategic equity partners. The Department of Posts & Telecommunications submitted a proposal not dissimilar to Telkom's but with a much shorter period of exclusivity and a stronger endorsement of competition.

The NTF, which met in November 1995 as the National Colloquium on Telecommunications Policy seemed unable to take a strong, coherent stand (Ministry of Posts, Telecommunications and Broadcasting, 1995b). The Colloquium's hoped-for "sufficient consen sus" always meant that the key players, Telkom (which was owned by the government) and organized labor (which was allied with the government), had to arrive at some accommodation for policy to move forward. Hence, notwithstanding the broad range of r esponses to the Green Paper which urged liberalization, the effective compromise had to include at least some period of exclusivity for Telkom in the public switched voice and data networks. How much time Telkom would have exclusive reign, and when the ne twork would open up in which market segments (and how such process could be properly managed) were the key market structure questions. How Telkom would access sufficient capital to roll out the network while becoming more efficient in preparation for comp etition was the other key question.

The Colloquium did and didn't arrive at sufficient consensus. A strong majority agreed that Telkom should have the primary role in universal service provision and there would be no competition with Telkom during a 3 to 5 year period of exclusivity. The t elecommunications infrastructure of other parastatals should be complementary, made available to Telkom under commercial conditions. Thereafter, the sector must be guided toward competition. The same majority backed Telkom's taking on strategic equity partners and/or private equity participation.

 

[rpm]

Labor continued to oppose the equity sale, supporting full state ownership and complete monopoly for Telkom indefinitely. These were "mandated" positions, laid out in the Green Paper submissions of Potwa and Numsa, and could not be changed with out further consultation. This development threatened to stymie further movement in the Colloquium. The solution was to kick the problem "upstairs." The question of market structure was referred to an Eminent Persons Group, a group of five " ;wise persons" to be nominated by the Colloquium Plenary to the Minister, and whose main task would be to oversee that the Technical Task Team embodied the letter and spirit of the Colloquium in the writing of the White Paper. The ownership question, and whether Telkom could take on private equity, was determined to be a general government policy issue with ramifications across other state assets, and was referred to Minister Jordan (Ministry of Posts, Telecommunications & Broadcasting, 1995c). T he specific question of Telkom's ownership was now recognized as just one part of a much larger debate taking place under the auspices of the Ministry for Public Enterprises on whether and how to restructure state-owned assets (Ministry for Public Enterpr ises, 1995).

Labor's concerns were, and are, not trivial. Parastatal reform does pose dangers. Among these is the retrenchment of labor, and in a country with over 40 percent unemployment this is no small matter. Labor also harbors concerns that its collective bargai ning position would be weakened consequent to any privatization. In addition, there is alway the danger that a privatized monopoly, given new freedoms and having to pay dividends to shareholders, will stray from its supposed developmental mission. As an i nfluential Cosatu-associated report argued, privatization may bring greater commercial efficiency, but it is typically at the price of service delivery and developmental priorities. Because capital always attempts to maximize return on its investment, Tel kom's private partner would surely push hard to increase tariffs, thus pricing the service beyond the ability of the disadvantaged to pay. Examining the privatizations in Eastern Europe, the report charged that schemes aimed at transfering ownership of st ate-owned corporations to disadvantaged communities have the consequence of leaving decision-making power with parastatal managers, investment funds, and stock brokers. Such schemes reduce state influence and enhance management's control (Makgetla, 1995). Similar criticisms were aired in the main publication of the South African Communist Party ( The African Communist , 1995).

Notwithstanding the dangers of reform, not reforming poses even greater dangers. This was the understanding of at least some of the Colloquium stakeholders who sided with the majority position. The abuses of monopoly providers in general, and Telkom in particular, were well understood. But politi cally motivated tariff schedules, poor service and excessive labor costs were only part of the problem. The monopoly carrier was simply not equipped to supply the needs of consumers in an era of rapid technological innovation. At the time that the cellula r market was liberalized, for example, Telkom provided service in a limited geographic area to 13,000. Eighteen months later, 400,000 subscribers to the MTN and Vodacom networks could use cell phones in all but the most remote regions.

 

[rpm]

7.2.1 The Government Plan: The Reconstruction and Development Programme

This position dovetailed with developments that had been taking place on another front. The RDP obliged Telkom to provide connections to schools and clinics -- an obligation deceptively larger than it appears. Given Telkom's financial difficulties, the c ompany forecast that it would need to borrow R6 billion over the next five years to fund its RDP obligations (Fleming, Martin, 1995). The company put out a request for proposal (RFP) in June, 1995 to provide one million lines in various rural areas where telephone density was below 5%. The government planned to work with two or more private partners, possibly chosen from the orginal equipment lisencees. The government insisted on full ownership for Telkom, effectively precluding project financing schemes such as Build-Operate-Transfer. Even so, the tender document stated that provision of the required infrastructure was to have no material effect on Telkom's debt/equity ratio (Telkom, 1995a).

Telkom learned some important things with the million line RFP. First, it learned that the arrangement could not be done without some kind of equity transfer. The notion that it could gain an operating partner without this altering the company's debt/equ ity ratio was fantasy. Second, Telkom learned that an RFP on this scale brought much better prices than it saw previously, on the order of 30 to 40 percent lower prices. With these two lessons in tow, Telkom opened up the process to a much bigger affair. In early November, Telkom unveiled a broad expansion plan -- which it called Vision 2000 -- of which the 1 million line tender was just a part. Scheduled over the next 5 years, the plan aims at increasing the network by 75 percent with the addition of 2 m illion lines in rural and urban underserved areas and 1 million lines in areas that already have well developed telecommunications infrastructure. Another 1.2 million lines would replace obsolete lines, aiming to complete the digitization of the network.

Telkom's normal rate of installation in recent years has been less than 200,000 new lines per year. The expansion of the network between 1994 and 1995 was at a rate of less than 5 percent (Telkom, 1995b). For the elections it was able to double that rate . (Telkom also services 800,000 transfers or disconnects per year -- a very high figure.) Because some of the Vision 2000 roll-out projects would be "turnkey" (that is, successful bidders would be responsible for end-to-end operation), particula rly in areas where telecoms expansion would be breaking new ground, Telkom's expectation is that it could install as many as 800,000 lines per year. This would amount to approximately a 15 percent rollout per year for five years. By way of comparison, net work rollout in Mexico was contracted at a minimum of 12 percent per year, a target which was met and even slightly exceeded ( Telmex Concession Document , 1990).

The financial terms of Telkom's Vision 2000 are sketchy and somewhat speculative. Telkom itself announced the cost -- at today's prices and without provision for price escalation, costs of financing, etc., at something in excess of R16 billion (Telkom, 1 995c). This is probably too low. It means that the average cost per line would be less than R4000, a figure never before seen in the South African context. The Department of Posts & Telecommunications estimated, conservatively, that the Vision 2000 pr oject would cost R20 billion -- and the Department averaged every service to the lowest figure (Department of Posts & Telecommunications, 1995). However, potential equipment suppliers confirmed that the R16 billion was a reasonable figure, in large pa rt because of the deployment of cheaper radio-based technologies in local loops (Hainebach, 1996; Botha, 1996). More recent Telkom estimates put the figure at R30 billion.

Regardless as to the most accurate number, this is a very large amount of capital to be raised and deployed in a very short period of time. Where will the capital come from? Telkom's current capital investment of about R2.4 billion per year, generated fr om internal funds, expands the network at a rate of less than 200,000 lines per year. If the big rollout requires an average of 800,000 lines per year, it is reasonable to assume from these numbers (admittedly on a simplified basis) that nearly three-quar ters of the new lines will have to be financed by new capital.

 

[rpm]

7.2.2 The Government Plan: The White Paper

Though Telkom's plans seemed optimistic, they were generally in line with the broad terms of the consensus reached at the Colloquium, and this consonance entered into the first draft of the White Paper. The White Paper articulated a market structure scen ario that gave Telkom essential exclusivity for five years, after which parts of the network would be opened up to phased entry and competition. specific timetable includes:

In year one, fully deregulating the equipment market and ensuring that cross-subsidies flow only to non-competitive activities (e.g. from business subscribers to low income subscribers).

In year three, issuing a third cellular license if appropriate.
In year four, rebalancing Telkom s tariffs to remove cross-subsidies; allowing resale, of excess capacity by private network carriers through interconnection with Telkom s network. All competitors would contribute to a "Universal Service Fund which would be used to fund network expansion.
In year six, allowing development of local loop service by local providers in cooperation with Telkom; liberalizing the long distance market; and allowing "metropolitan area networks," or networks of private networks providing service to major cities.
In year seven, licensing a second network provider; and liberalizing the international call market.
The White Paper market structure scenario reflects the particularities of the South African situation, and is at the same time reasonably parallel to reforms undertaken in other countries. It gives the public telecommunications operator a period of time to expand even as it facilitates the growth of value-added and private network entrants, and phases in regulated competition. In this respect the plan commits the system toward a massive network roll-out while also permitting the development of sophistica ted value-added services through the phased regulation of competition -- first at the margins, then at the core of the infrastructure. And the massive expansion of the network may alleviate the problem of potential labor retrenchments, inasmuch as network expansion will require considerable manpower.
However, its plausibility is contingent upon two crucial things. First, the government must establish a competent regulatory authority very quickly, for the Regulator will be expected to oversee many complicated processes and resolve a great number of te chnically intricate disputes. Guiding a sector into a competitive market structure is no easy task. Regulatory disputes are very likely to be dominated by Telkom, particularly at the beginning, inasmuch as it possesses most of the key information. The Reg ulator will have the difficult task of monitoring Telkom's activities to distinguish between actions that facilitate the central goals of the sector in terms of reconstruction and development, and actions whose effect will position Telkom so powerfully as to undermine eventual competitors. The White Paper proposes the creation of a Universal Service Agency alongside the Regulator, to ensure that universal service remains a central focus. It also may be advisable to write as many obligations and understand ings into Telkom's license as possible, particularly because it will take time to built capacity in the Regulator. Performance and regulatory contracts have worked in some countries, Mexico being perhaps the best example. These would provide needed initia l stability for the new market structure and at the same time afford the regulatory body time to build up its capacities.

Second, assuming, conservatively, that Vision 2000 costs R16 billion, Telkom will need another R12 billion, or R2.4 billion per year, for the plan to work. Telkom cannot effectively expand considering its present financial state. Telkom is the most highl y "geared" (net debt as a percentage of shareholder's funds) of any telecommunications company in the world. Its capital expenditure is below other comparably situated national telecommunications operators, and this is reflected in the low rates of network growth (Telkom, 1995b). Without new capital, Telkom cannot expand the network in anything approaching the rapidity envisioned by the White Paper -- a speed both expected by the South African majority and necessary if Telkom is to meet the chal lenge of competition from international operators.

 

[rpm]

7.3 FINANCING REFORM

There may be several ways to infuse capital into a debt-burdened Telkom. Government funding is not an option at this time given the severe budget constraints facing the Government of National Unity. General sale of equity on the Johannesburg Stock Exchan ge is a possible option, but such a gambit risks replaying the consequences of the previous government's strategy because it would tend to concentrate ownership in the white elite. A give-away of shares to favored, previously disadvantaged constituencies would increase social equity, but would not raise capital for Telkom. Although the million line tender has prohibited this option, build, own, and transfer (BOT) arrangements, which grant large, usually international contractors an exclusive right to a pr oject's revenues under a time-bound concession agreement with the government, have worked in other countries, including, apparently, South Africa in the case of Water Sanitation Services South Africa's arrangement with the French company, Lyonnaise des Ea ux. However, governments are not always able to manage the impact of levies or tariffs on the public.

In early 1996, the government enlisted an investment bank, Goldman Sachs, to help it find a strategic equity partner. A strategic equity partner (SEP) was considered the most propitious way to facilitate an infusion of capital into Telkom. Not only would this raise capital quickly, but the participation of an international telecommunications operator would likely bring much-needed management and technological capabilities, as well. The book value of Telkom's investments, including buildings, etc., is onl y about R15 billion, but the parastatal's potential market value is estimated between R31 and 35 billion (Fleming, Martin, 1995). Telkom's Chairman of the Board Digang Moseneke has consistently been quoted in the press as looking for a strategic equity pa rtner of between 20 to 30 percent. If Telkom's market value is, for purposes of argument, R30 billion, an SEP would represent a direct infusion of capital between R6 and R9 billion. Telkom would probably also borrow new money on the local and internationa l capital markets, hoping to leverage the SEP's contribution toward a favorable interest rate. Even so, it seems that even with the SEP, Telkom will have difficulty funding its obligations under the RDP and Vision 2000 and becoming a viable competitor wit hin seven years. The "showdown" between the ANC and its labor ally over the issue of Telkom taking on an equity partner was expected to come to a head in mid-1996.

 

[rpm]

7.4 THE FUTURE

Continued Sectoral reform requires the government and organized labor to come to a reasonable and rapid settlement of the political problem on the restructuring of state assets, a debate which pits the principal coalition partners of the ANC and Cosatu a gainst each other. The White Paper itself specifically leaves the question of ownership to future discussions, although by the end of 1995, the government had begun discussions with several international firms. In January 1996, Cosatu announced its intent ion to hold a general strike in response to the impending partial privatization. Fearing the economic and political consequences, the government agreed to delay the sale of equity during talks which resulted in The National Framework Agreement. The agreem ent permitted labor a voice in the restructuring process and some degree of control in any layoffs which would result. In exchange, Cosatu dropped its unconditional opposition to privatization, though its continued to voice its preference for state contro l of its parastatals (National Framework Agreement, 1996).

The telecommunications reform process was thrown into uncertainty due to the dismissal of Minister Pallo Jordan from the Cabinet, effective April 4, 1996. The Cabinet reshuffle also dismantled the RDP, its functions to be distributed to line ministries. Jay Naidoo, head of the RDP, was given Jordan's Posts, Telecommunications and Broadcasting portfolio. Speculation on Jordan's dismissal rested not on job performance so much as his maverick style and the fact that he had crossed President Mandela on sever al occasions.

Potential investors were initially alarmed at the elevation of the former labor leader and opponent of liberalization to a position responsible for the future of the telecommunications industry. Naidoo issued several statements proclaiming his committmen t to the reform process, and analysts began to speculate that he might actually command a better bargaining postion from his former collegues than the intellectual and apparently aloof Jordan. Draft legislation issued in May, 1996, suggested that Naidoo w ould continue in the spirit of the NTF and the White Paper. The bill, due to be debated during the 1996 parlimentary session, further delinated distinctions between policy, regulation and operations by creating the Regulator (the South African Telecommuni cations Regulatory Authority or SATRA) and the Universal Service Agency, and consoldating policy authority under the renamed Department of Communications.