The Impact of the Internet on Traditional Telecom Operators

Introduction

For more than a century, telecommunications around the world followed a classic model: a national monopoly owned or controlled by the state, centrally managed and providing a common public network. By their very nature and tradition, these networks provide a small number of standardized and nationwide services, carefully planned, methodically executed, and universally distributed. But over the past two decades, first in the United States and subsequently in much of the developed world, the forces of centrifugalism began to unravel this traditional system. The driving force behind the restructuring of telecommunications was the shift toward an information‑based economy, which resulted in the rapid growth and reliability of telecommunications as the medium for the electronic transmission of information. Especially for large organizations, the price, control, security, and reliability of telecommunications became variables requiring organized attention. In a series of steps, each controversial and painful, monopoly began to give way to the "network of networks."

The conventional scenario for the evolution of telecommunications, offered in the past by traditional telecom operators as their vision of the future, had been the integrated single superpipe, merging all communications infrastructure into a single conduit controlled by themselves and interconnected internationally with similar territorially exclusive superpipes. This scenario of integration took no account of the simultaneous organizational centrifugalism that was taking place. Instead of consolidating, the network environment kept diversifying horizontally as new carriers entered. This has been noted and commented upon. But now, there has also been a trend towards vertical diversification. On top of the physical carriers, there are resellers, private networks, integrators, and Internet-based service providers.

The Impact of the Internet on the Structure of the Public Telephone Operator

There are several fundamental ways in which packet-based IP communications differ from traditional circuit switched telecommunications and affect it.

• Carrier costs are lower in an Internet system since they can shift many intelligent functions to the users and away from more expensive switching intelligence.

• Internet transport services are more homogenous than switched services and this leads to the commoditization of transport.

• Pricing of IP services tends to be market and costs based, whereas the prices of traditional telecommunications providers were traditionally forced down or up by regulation.

• The relative decline of voice as a share of traffic, relative to data, is accelerated by the Internet. Voice will lag because there are limits to the human time and to the brain’s information-density, whereas no such limit exists for machines. There are also many more machines than people, and machine-based communication will move down from computers to numerous smart devices. Refrigerators might be able to order groceries on their own, after shopping for the best deals. Cars will communicate with roads and other cars. Data bits are cheaper than voice bits because they need not be in synchronous real-time. Together, these factors lead to an absolute and relative increase in data traffic, which means a decline in the importance of circuit switched architecture.

• Internet applications increase demand and utilization. In particular, the introduction of individualized video entertainment on the Internet will increase capacity consumption enormously.

Threats to the PTOs from the Internet

• Competition for bandwidth. More than half of the traffic crossing between the US and Japan is data.

• New competitors. Because IP is easier to provide than switched voice, new carriers such as cable television companies, wireless operators (narrow and broadband), direct broadcast satellite companies, electric utilities, and others will compete with telephone companies in providing IP networks.

• Internet Telephony. An IP-based infrastructure can be used for the diversion of voice traffic from the PSTN. Internet telephony has succeeded in overcoming its initial barriers of low quality, non-ubiquitous hardware, and lack of interconnection. Today, Internet telephony is approaching wireline voice quality. The industry itself, which was once based on non-standardized shareware software applications, is homogenizing. For example, the ITU has issued the standard H.323 for Internet telephony. One of the largest threats looming on the horizon is the ITXC, or Internet Telephony Exchange, which provides accounting services and allows Internet Service Providers to share each others infrastructure resources.

• Internet Fax. Because fax does not require real-time transmission, the Internet can provide identical services. Users can provide their own broadcast fax facilities, store-and-forward fax systems, receive when busy capabilities, plus others.

• Internet as full Service Provider As data moves from a secondary telecommunications service to the primary one, the Internet might be the opening which will lead customers away from traditional carriers, providing telephony through an IP based service or through leased or owned switched networks. (So far, however, American Internet providers have been are too fragmented and unstable in quality (e.g. AOL) to provide any strong national brand identity.)

Opportunities for PTOs Created by the Internet

At the same time, the Internet also creates significant opportunities for traditional telecommunications operators.

• Provision high capacity transport services for national backbones and local ISP operators. AOL alone plans to spend 1.2 billion dollars in telecommunications expenses.

• Dial-up users stay on the line for long periods, leading to increased revenues.[1]

• Users often migrate to higher speed facilities to connect to the Internet, such as ISDN and T1/T3 services.

• Second lines for the Internet. Approximately 1/5 of households in the U.S. now have second lines.

• The provision of Internet services by traditional carriers.

• The provision of service elements for other Internet providers. As an existing hub for switched traffic, carriers can serve as central facilities for Internet-based activities. This could include both web and server hosting.

• The provision of extranets and intranets to large customers, and managing their integration with existing internal services.

• The use of IP-based technologies for operation of regular services such as voice at a lower cost.

• The provision of backbone services. MCI/Worldcom marketshare of backbone revenue is over 60%, and Sprint is the second largest carrier.

• Metering, billing, network management, and directory services for the ISPs.

• The provision of integration services such as interconnection points and clearinghouses, gateways between the Internet and the PSTN for Internet telephony, gateways between the Internet and other data networks/services, domain name server management, and shared points of presence facilities.

The pressures of liberalization are forcing carriers to choose whether to embrace this new Internet world of greater opportunity, fight it, or ignore it. But it is unlikely that the status quo will remain for long.

Impact on the PTO Organization

• Fragmentation by Geography. These factors have consequences. The pressures of vertical and horizontal diversification are likely to force a fragmentation of carriers. As carriers are forced to compete on costs for multiple services, geographic service areas which lose money are likely to be spun off or be outsourced. In America, US West has been following such a policy. If such divestiture is not possible due to “carrier of last resort” obligations, telecom firms may segment themselves into universal service companies and competitive units.

• Fragmentation by Conduit Type. Carriers may be fractionalized by conduit type, e.g. wireline and wireless networks.

• Fragmentation by Network Elements. As networks must provide interconnection in an unbundled fashion, each element of the network compete on its own merits. This creates centrifugal forces in the company structure.

• Fragmentation by Customer Class. Carriers will fracture along business lines, such as business and residential customers, based on competitiveness, customer sophistication, quality levels, and price elasticity.

In the Internet industry, service providers tend to target either business (UUNet, BBN) or residential (AOL, Mindspring) customers. For example, PSINet divested itself of all its residential customers to concentrate on the business market.

• Fragmentation by Capital Markets. Competition and Internet entry will necessitate new infrastructure development, while at the same time revenues will be under pressure, and risk is increasing. Under scrutiny from investors, unprofitable units are dropped, and diversified companies -- “mixed-plays” -- are discouraged due to the greater difficulty in analyzing their value and the greater problems in acquiring or selling them in a merger. An example is the separation of Lucent from AT&T, which helped the share prices of both companies.

• Fragmentation by Regulation. Regulation served to prevent the fragmentation of carriers by requiring operators to cross subsidize between different categories of customers and services. Liberalization permits unregulated carriers that specialize. Faced with competition in some markets, and the potential for cross-subsidies to the competing segments from the monopolistic ones, regulators have required various separation arrangements (divestiture; fully separated subsidiaries; accounting separation; etc.)

• Fragmentation by Source. Internet and other services may be separated. This will be discussed further below in the section on corporate culture.

Structurally, the major division within the firm’s carrier network operation is between local and long distance. What is it for Internet provision? A similar division would apply. ISP--Internet Source Providers-- is an imprecise term. It is applied first, for the service provided to customers on the local level, providing connectivity to the Internet after being reached through a direct or dial-up phone connection. This kind of service is provided, in the US, by about 5,000 firms, most of them small. These local retail Internet Access Providers connect, directly or indirectly, into National Backbone Providers. In the US, there are about 40 of these companies. They operate mostly over leased lines, usually refuted from long distance telephone companies, and connect into each other at various network access points and through direct peering traffic hand-offs. These backbone providers resemble long distance telecom firms. In some cases, the IAPs and the backbones are vertically integrated, but this does not change the basic distinction.

Thus, there are four major functions to network and Internet provisions: local; long-distance transport; Internet access; national backbones. There are advantages and disadvantages to their integration. Among the advantages are economies of scale and scope; sunk cost; demand externalities; one-stop shopping; coordination and control; price discrimination; product differentiation; transaction costs; market power.[2] And others. On the other hand, the specialized firm has the advantage of focus, faster product cycles; and the ability to do business with more firms without conflicts.

For these reasons, telecom firms will need to re-organize to meet the challenges of the Internet. Several options exist, outside of the status quo:

Strategic Options for Structure

1. The Wholesale Network

PTOs can specialize in the network part of their operations, selling capacity and special elements to providers of final services. As the Internet grows, this wholesale business becomes evermore important for traditional carriers. Many of the CLECs in the California region are wholesaling xDSL services to ISP who then resell it to their customers. New Internet-oriented carriers such as Quest and Level 3 and even the satellite carriers like Teledesic base their entire business plans on wholesale models.

For Internet services wholesale markets can occurs at multiple levels. There are also elements outside the transmission which could be unbundled. These include: billing and collection services; customer support networks; and network maintenance.

2. Resale of Services

While it might seem unlikely that a traditional network operator would switch to become a reseller, it is simply a question of economics. If West, using a low-cost, high capacity architecture, is able to wholesale for less than 1 cent per minute and AT&T's costs are estimated at 1 and a half cents per minute, AT&T would be making a poor economic decision (all other things being equal) if it did not resell West's service.

To the carrier’s advantage, a nationally recognized brand name provides an important asset in a resale model. This is apparent in the large number of telephone customers in the United States who still believe that their local telephone company is still AT&T! With Internet resale, services can be seamlessly resold to consumers who have little knowledge of the underlying carriers.

3. Systems Integration

The consequence of such an interconnected and unbundled network of network is the emergence of systems integration as a key institution of the telecommunications of the future. This is what Internet Service Providers offer: systems integration, with a bit of their own hardware added.

Systems integrators assemble packages of modules of hardware and software and structure these packages to the needs of their customers. To these customers, the identity of the underlying carriers and their technology is unknown and transparent.

Systems integrators might operate a least-cost-routing system, switching users around as capacity becomes available. They can function as capacity brokers, buying and selling capacity as it becomes available. Likely to emerge is an international market in capacity, consisting of a futures capacity market and a spot market operating in real time.

To be successful as a system integrator, carriers must be willing to ruthlessly pick and choose modules, if they compete and drop their own companies' modules if they are not better and cheaper than other modules. Similarly , the underlying carriers cannot favor their own systems operators.

There will be separate markets for separate network elements. Some markets will be wildly competitive. Others probably may be monopolistic. Pure trunk transmissions modules will probably be basically a commodity business. The systems integrators buy transmissions elements at marginal cost, and make it a successful business, even if underlying carriers cannot.

This is not mere hypothesis. In the US, Pacific Telesis reorganized itself in 1994. A voluntary self-divestiture, when it spun off its mobile subsidiary. The new company, Airtouch, has to negotiate a fair interconnection contract with Telesis free of "fraternal" preferences. Telesis did not give Airtouch any preferences. Airtouch went to other carriers for carriage. In time, PacTel began to reenter mobile communications, in competition to Airtouch.

The Rochester Telephone Co., a medium-sized independent telephone company in NY, separated in itself into a wholesale operation offering transmission to all, including its competitors, as well as a retail services operation.

Outsourcing and the Virtual Operator

Systems integration, taken to its extreme, leads to a “virtual organization”. That is, all functions of the telecommunications firm are acquired from other suppliers. The role of the company is to assemble, not to produce. Internet Service Providers outsource all or most of the transmission functions. But the outsourcing need not stop with transport. Outsourcing or provisioning may also include other functions: marketing; billing and collection; research and development; customer relations; government relations; etc. Even the labor force can be provided by contract firms, while the middle management may be provided by consulting arrangements with free lancers or management firms. In the extreme, the firm would consist of no more than its top management, and may not even have a physical headquarters, substituting instead an intranet.

The economic advantages of such an arrangement are that the firm can avoid fixed costs, and transform them into variable costs. This lowers the cost of entry while raising its speed. A firm can thus also benefit from specialized providers experience and economies of scale. It can acquire these inputs competitively, contracting with the optimal combination of price and performance. It can shift much of the risk to the outside providers. It can acquire inputs produced offshore in low-wage, low-protective countries, or under non-unionized conditions, or bypassing various forms of regulation.

On the negative side, the firm becomes dependent on others for critical inputs. It loses the synergies that may come from combining production with application. It may contribute, through lowering the cost of producing the input, to lower costs for its competitors, too. It cannot establish a loyal workforce in whose skills it invests. It may not have a corporate culture or institutional memory that would contribute to its operation.

All this suggests that total outsourcing would be unlikely. This might suggest that the firm self-provides its core operations, while buying peripheral inputs elsewhere. Indeed, few firms produce the buildings or the automobiles they use. And almost no telecommunications firm is still in the business of equipment self-supply, after AT&T spun off Lucent. However, the advantage of outsourcing could go considerably further than in the past–partly because information technology makes the coordination of such an operation more possible. These advantages would be exercised most actively by new entrants who have not established their production facilities. In comparison, it would be more difficult for an established company to exit from self-provisioning.

They would use their own inputs, produced with a considerable fixed cost element that would need to be recovered, whereas their rivals, in a competitive environment, might be able to buy these inputs at very low marginal costs, in the case of transmission.

To add to the problem for the established firms, they will often be required to supply the inputs to their Internet rivals under regulated conditions. This means unbundled elements; pricing according to long range incremental costs; and quality requirements. Yet their competitors would not be under simultaneous obligations. Hence, when the entrant had a cost advantage, it would be left with it. But if the incumbent had such an advantage, it would have to be shared with its rivals. This asymmetry creates competitive disadvantages for the incumbent.

In consequence, there would be strong pressures for the incumbent to operate, too, as an outsourcer. And this means that it would split its operations into an integration operation, i.e., a service company, and into input production, i.e., a network operation. Or, several service companies, especially one for Internet service, and several network and input producers, including those offering transmission capacity to ISP’s.

These entities would deal with each other partly in a coordinated fashion, under the direction of central management. But they are likely to strive to obtain the best terms in the market, in order to stay competitive in both input and output markets The market prices would also solve the problem of how to price intra-firm transactions, and how to compensate for the risk which one part of a parent company might impose on another.

Provisioning can become a profit center for the company. For example, the billing and collection system is a major asset which could be utilized in additional ways. It could be offered even to Internet competitors. It could be offered out-of region, or internationally. But it must also be recognized that distance -insensitivity makes other telecom companies, such as those from other countries, also potential rival input suppliers. This means that cost and performance pressures will be strong.

Even more importantly, the local loop can become a major outsourcing profit center. As competition tries to emerge in local service in the US, Japan, and Europe, it becomes clear how expensive it is to provide rival residential local loops. Cost is high and competitive entry is difficult. In the future, for an international call, the local transmission segment may be more expensive than the international one! Thus, the provisioning of this local segment is likely to be the main source of competitiveness of telecom organizations.

Outsourcing permits a traditional firm, too, to rapidly enter a new market. An example is video programming, whereit does not have major expertise, but where it could outsource to other firms that specialized in channel provision.

Markets for Outsourcing Elements

In the past, telecommunications bandwidth was a relatively static resource. It was centrally planned and controlled by the established telecommunications operators. Its retailing to endusers was almost always undifferentiated as to price and performance. Such narrowband capacity was distributed fairly uniformly across the country through such policies as universal service. The key goal was widening the distribution of narrowband capacity across geography and society.

Now, much of this is changing. Now the deepening of capacity is the focus. Differentiation rather than uniformity is often the strategy. This changes the nature of provision. Whereas only a few years ago, the expectation was that increased capacity would be offered by the same system, simply with bigger pipes, the emerging system is much more complex.

The first factor of change has been the liberalization of telecommunications services and the emergence of competition. This created numerous parallel conduits of bandwidth. This, in turn, accelerated the resale of capacity, and therefore led to the beginnings of capacity markets. The second factor is change in traffic patterns due to Internet usage. Telecom networks have moved beyond 2-way communications to support more sophisticated “push” distribution models, and demands for transmission vary greatly across users. The third factor is the trend towards packet networks. In a packet switched network, information can travel over multiple pathways to its destination. Instead of reserving identifiable capacity, users now contribute their transmission needs into the “cloud” of the network of networks. This resembles the pattern for electric grids -- continuous connectivity, pooled provision, and unidentifiable origin of who produced of the units actually consumed.

As production and consumption of bandwidth become more decentralized, fragmented, and unpredictable, a new market system for capacity will thus emerge. Just as modern financial institutions aggregate capital and provide for the creation and expansion of business, so will bandwidth markets evolve to provide the bandwidth required to meet the growing needs of international information industries. Market institutions will be created to establish prices and assure fulfillment between conduit providers, third-party brokers, and end users. In that market, carriers will offer capacity, as will some users. Buyers will be endusers and intermediaries such

as other carriers and systems integrators. This transition to a capacity market will therefore be the next step in the evolution of networks.

How to account for returns?

If the telecom organization offers several traditional as well as new style Internet services, the question is, how to allocate costs, revenues, and profits among the different activities of the company. This question is important for regulatory purposes, if part of the enterprise is subject to price or profit regulation. In such a case, the company will try to shift costs into the constrained segments, while shifting revenues into the unconstrained one. The question is similarly important for the prices that the company charge to its rivals for access, interconnection, and wholesale services. And it is also important inside the company in judging performance of the sub-units.

For these and similar reasons, the attribution of costs and revenues is one of the critical issues in telecommunications management and regulation.

When it comes to the Internet, the allocation questions abound . The company leases lines to ISPs as an intermediate input. It also provides access to the enduser from the ISP. And it competes against the ISP in Internet and telephony service. The incentive is to charge its ISP competitors as high a price as possible, while keeping similar charges to its own Internet operations low, in order to keep its Internet prices low. The more important the Internet operations become, the greater the incentive to engage in such profit shifting to the basic carrier operation. In so doing, the carrier operations are likely to have the support of the traditional regulator as long as these profits are used to keep residential rates low.

But, as the number of Internet users increases, they become a mass constituency with influence in the regulatory and political process. Their interest lies in cheap and high performance local facilities, and in cheap rates intermediate inputs for ISPs. The two goals, practically speaking, are in conflict. Generally, Internet users in America have chosen to support low ISP prices over low residential rates. Thus, they have been opposed to the ISPs paying access charges to maintain universal service, which is a major source of revenue for local exchange companies. Squeezed from both ends, the local companies seek as source for redistributory payments the traditional cash cow for telecommunications, the large corporate users. However, these users now have options from other carriers, and migration to the Internet is one such option

The main way to account for revenues is through two steps. 1. The segmentation of the network operations into multiple elements. 2. The establishment of competitive markets in all segments. Together, these policies create market prices in the use of the elements. These prices can then be used to allocate revenues. Suppose that the second condition is not met, and that markets are incomplete or inefficient. In that case, the alternative is to engage in some estimation, using some proxy mechanism on how the prices might be in a hypothetical market. One way to do so is through the creation of a hypothetical engineering model of a network, and using these prices, with some mark up for a profit, as the price for the service. This is, basically, the approach taken in the USA for the pricing of interconnection. The methodology is known as TELRIC, supplemented by complex computerized cost models.

If the first condition is not met and the company produces a variety of services within the same entity, the engineering cost methodology can still be used. However, the major problem in that case is the allocation of the fixed costs, as well as the historic costs that may include some expensive equipment that would not be used today anymore, and which has not been fully depreciated.

The Corporate Culture of Telecom Organizations in the Age of the Internet

Technological change leads to business changes, which in turn require cultural change. Corporate culture is the “ pattern of shared and stable beliefs and values that are developed within a company across time” (Gordon and DiTomaso, 1992)[3]. They are based on commonality: shared history; shared values, shared goals, shared leadership, shared processes, and shared interest. Some of the cultural attributes can be mandated from the top. IBM, for example, for decades required its male employees to wear white shirts. But much of corporate culture needs no formal mandate and is instead a behavior and value pattern grown over time. In most Wall Street investment banks it would be unthinkable for managers not to wear a tie on weekdays. On a weekend, on the other hand, a tie might not be appropriate. But in most cases, culture is not handed down from the top, but an integral part of the organization.

For more than a century, telecom organizations operated with a culture shaped by engineering and civil service value systems and operations: clear and specified procedures; clear lines of responsibility; long planning horizon; job security; a politicized decision making; a public service orientation; a national and social perspective; risk avoidance; and a management that rose slowly inside the organization, having adapted to its values..... In America, those that shared the dominant telco culture were known as having “Bell Shaped Heads”. Other countries had their similar types.

This traditional telcoculture cannot survive the simultaneous challenges of privatization, competitive markets, globalization, and convergence. All of these are embodied in the most potent cultural challenge, that of the Internet.

The Internet culture draws from other well springs: entrepreneurialism; individualism, informality, risk taking, rapid product cycles; uncertainty, informality, and risk-taking.

These two cultures are now clashing as their industries compete, cooperate, and overlap. Will the result resemble the cultural transformation of Japan, which radically transformed itself in the Meiji period under the influence of the West? Will it be more like Peru, whose indigenous culture was destroyed by outside force? Or will it be more like Byzantium, assimilating diverse cultures into a new whole?

The strongest motivator of culture is success. An effective culture makes for an economically successful organization (Kotter &Heskett, 1992), and success reinforces a corporate culture. This means that the dominant telco, confusing strength with effectiveness, (acquired under the protective umbrella of the monopoly status) believes itself to be successful, and hence is resistant to change.

Yet even where management embraces cultural change, corporate culture is much slower to change than organizational structure, top leadership, or strategy. All of these can be changed by rapidly by decision. The collective values and the way people do business change much more slowly, because it is the aggregate of many behaviors and routines acquired over a lifetime.

This means that the culture is likely to be a drag on an organization in change. An organization with old values has to confront competitors who have started with the new culture proper to the task.

How then should traditional telecom organizations respond? Northcote Parkinson, the noted English guru of organizational pathology, advocates in one of his laws the total replacement of every last employee. For him, even a few hold-over apples will spoil the barrel. But such a policy seems neither practical, nor legal, nor consciounable.

A related strategy is to accelerate the turnover of employees, by offering early retirements, and shocking the remaining employees into adjustment. But this can be a costly proposition, directly in retirement benefits, and indirectly by the potential loss of the most productive employees who leave to follow ouside opportunities. There will also be an additional loss in employee morale. This option would be less likely in Europe, where downsizing is more difficult.

The complementary strategy to a firing of the old is the hiring of the new. But bringing in new people with supposedly different values creates problems, too. If they are young, they may be malleable and powerless, and their values may soon conform rapidly to the existing orthodoxy, while reducing their job satisfaction. If the new people, on the other hand, are hired at the middle or high levels of the organization, they will likely clash with the existing managers who will exult in their failures. Thus, fractionalization and in-fighting are likely.

Under the best of circumstances, some amalgamated new common culture may emerge that forms an internal equilibrium. But that might not be appropriate neither to meeting new challenges nor old ones. A style combining that of the Internet and of traditional PTOs is likely to be unsuccessful in neither market, while being stressful to both parts of the organization.

Another approach is the re-socialization of the existing work force. Here, top management tries to redo their own subalterns in their own image. In America, an entire industry of specialized consultants has sprung up to accomplish that purpose. But like similar attempts of re-education in China under Mao, these efforts are doomed to failure, producing at best hypocrisy and at worst obstructionism. This may be counter-productive as even those aspects of the old culture that deserve respect, such as the public service orientation of telcos, are denigrated. Those managers most adept in mouthing the slogan of the day get promoted. And form becomes valued over substance.

Much more effective than organizational sensitivity training is to provide the proper incentives and organizational structure. Such a structure establishes behavior and processes, which in turn affects the performance.

Therefore, the best structural strategy of cultural change is corporate multi-culturalism. This means the segregation and coexistence of different sub-cultures within separate business units. In practical terms, it means that the traditional organization creates autonomous units whose culture and style can markedly vary from the traditional one. If successful in its realm of activity, the new culture will reinforce itself in its own unit, and might spread to other parts of the organization, this time with the legitimacy of success and in-house origination. And if they fail, the harm is contained mainly to the sub-unit, not to the entire organization.

Such a strategy can be fairly readily adopted for new lines of business. Mobile communications, for example, can be organized as new and different organizations, and this indeed has often happened successfully. Internet operations can similarly be modeled on ISPs rather than PTOs.

But how can the change be spread to the core of the operation, traditional telephony? One way would be to decentralize the operation, by region, line of business, service, etc. The regionalization of the Bell companies and the separation of long distance from local operations in America resulted in increasingly different corporate styles. Overall, American telecommunications benefitted, despite the greater untidiness. It became clear how despised the style of the AT&T headquarters--195 Broadway in Manhattan--had been in the live organizations. Similar changes might be in store in Japan. In both cases, the decentralization took place under coercion of the government. The much preferable approach would be for the PTO to reorganize itself, both for practical operational reasons, and also in order to permit different cultures to emerge, and to contest each other inside the organization.

Corporate leaders, instead of trying to put a “holistic” cultural stamp on the organization as a whole, need to give substantial cultural autonomy to the sub-units. That means acceptance of different styles, values, incentive systems, compensation, and promotion systems. It means the acceptance of a weakened center. It means the acceptance of rival goals inside the larger organization, not only tolerated as unavoidable, but instead accepted as beneficial.

In the new environment, the notion of shared organizational history has already become largely irrelevant. To the Internet part of the organization, history starts in the 1970s, not the 1870s. Similarly, the identifiers of a common entity, such as the corporate logo, color, anthem, etc become useless as an integrative device.

Whether the leadership will like it or not, the shared values will not be shared anymore. Different parts of the organization do things differently, and for different reasons, and with a different vision of their future. Some empower individuals, others control them firmly. Some take a short term orientation, others look for the long term. Some provide risk and incentives of wealth, others the enrichment of an interesting environment, others job security. Some reflect domestic values, others are shaped by those of other home countries. Some are hierarchical, others have a flat pyramid as an organizational structure.

In such an environment, what is the role of shared top management? Is it necessary? Is it possible? With operational and cultural autonomy with the sub-units, central management becomes, essentially, a holding company of existing, newly formed, and acquired companies, both domestically and abroad. In some cases, the centrifugalism might be too strong for the company to hold together. Airtouch, for example, was the entrepreneurial part of the traditional Pacific Telesis; and was spun out of it as an independent company.

In some ways, this structure might be more easy to adapt in America than in some European countries, because in America the culture of diversity has become dominant. The Internet culture fits into broader changes in society. And management can take more radical steps with some impunity. European countries operate with more constraints. They are also less likely to imitate each other more than American firms. This makes it especially important for leadership to emerge inside a European telecom organization. Italy, with its tradition of decentralized operations, can teake a lead.

The Impact of the Internet on the Regulation of the Traditional PTO

It would be nice to imagine that the Internet future is a future with minimal government interventions. But that is not likely, since the Internet, with its wide scope of activities, will mirror the complexity of the underlying society and economy, and will reflect the rules set outside of telecommunications, which tend to be substantial.

Because the Internet is a force of change in telecommunications services and networks, it will also affect the way that governments regulate telecommunications. Those changes are intertwined with the more general ones associated with privatization, liberalization, competition, and globalization. However, there are distinct, Internet-related issues of regulation that one should anticipate. They affect the type of constraints and opportunities that one can expect for the future.

Traditional Universal Service Financing

Internet-telephony, especially international ones, contribute rapidly to the undermining of the traditional system of high international rates that support low local and subscription prices. The problem will be both on the revenue side and on the cost allocation side. On revenues, the prices of many of the services that were contributory to the universal service subsidy will collapse. Other sources of revenue must be found, or the subsidy must be reduced, or the two approaches be combined.

On the cost allocation side, the change will go beyond the conventional “rebalancing” of prices to match cost. The issue is the more complex allocation of the fixed costs. And here, the optimal allocation is by the Ramsey “inverse-elasticity” rule. This means that those services with the least elastic demand will be disproportionately burdened. In contrast, the elastic portions are likely to be the long distance and international segments of networks. This means a shift of cost onto the local portions of networks, a reversal of past arrangements. The local network will hence be the key revenue source for operators and universal service.

Internet telephony will be integrated into the payment mechanism for universal service. At the same time, the resistance of the Internet community to such requirements will lead to a revision of the traditional universal service system and to its narrowing to the poor and the rural population instead of also subsidizing the broad middle class.

In America, the FCC, in a report to Congress on April 10, 1998, already cautiously suggested that phone-to-phone Internet telephony be required to pay the same access charges as the long distance carriers. These access charges have historically been above cost and contributory to local and high cost operations of LECs. The Internet community has greatly opposed having to support inefficient operations of LECs. They also argue, make an “industrial policy” argument, that any charge would impede the growth of the new medium, with its overall beneficial impact on innovation and the economy.

Treating Internet telephony differently from other forms of long distance service creates a major non-neutrality among competitors. Because the access charges into the local networks are quite a substantial burden, the IXCs are beginning to shift their operations to an IP model due to its favored regulatory treatment. This would be inefficient if it is based on regulatory reasons rather than technology and operational economies. Furthermore, the distinction would require continuous monitoring and controls to ascertain that a supposedly IP-based service indeed uses IP, as it claims. It really should make no difference to the regulatory treatment what technology or medium a long distance service uses to move its bits.

The basic problem is that ISPs, and with it Internet telephony is considered a use (not requiring to pay access charges) rather than a carrier. This distinction is ultimately futile and leads only to major regulatory headaches. The proper treatment is to charge equally for the use f the segments of the local networks, based on the actual incremental cost, without a subsidy element.

To cover the fixed costs, the likely resolution is a flat charge on local access lines. The FCC has already done so in the past through subscriber line charges (SLC). It has recently continued this policy through a charge on the long distance carrier for each pre-subscribed customer, which is similar, though less visible than a SLC, and less useful for IP telephony access without subscription, because lines and subscriptions are likely to create definitional problems in the future. More likely is therefore a charge based on phone numbers and capacity.

A second category of charges supportive of universal service is a type of a tax --excise, sales, or value-added--charge that all carriers pay into a universal service fund in proportion to their revenues. This system to support universal service seems a logical and simple system for the competitive environment. But here, too, the question is whether competitve IP providers will be treated differently.In the U.S., the existing system taxes all carrier revenues, not just those for transmission. By including the services, too, the question of how to treat the Internet service providers is raised.But, such a question would not be relevant if only transmission were taxed, and not the service. With such a system, IP telephony providers would not be taxed directly, and would be burdened only indirectly, like everyone else, through their use of transmission services from carriers, which would be taxed. But how could one separate transmission revenues from service revenues? By their organizational separation. Such a separation would not be mandated. But the incentive to do it is strong because it reduces the tax burden. If the company tries to shift revenues away from transmission to services, it will invite competitors to buy the transmission at the artificially low price.

Thus, the integration of IP telephony will require a reform of the financing system of the Internet: away from usage sensitive and to flat rate pricing that is cost based; and away from a general revenue tax to a transmission revenue tax; and to a separation of transmission operations from services.

These changes are independent of the size of the universal service subsidy. They would be equally applicable however large or small the subsidy is (unless it is zero), or who its recipients are. However, an additional and important question in what direction the universal service subsidies will trend: will they increase or decline? Many people expect that competition will lead to greater operational efficiency and hence a smaller needed subsidy, all other things being equal. But this is wishful thinking. To the contrary, the size of the universal service subsidy is likely to increase substantially, as the importance of telecommunications connectivity to participation in the community, society, and economy increases dramatically. And as new technology emerges and applications abound, the demand for participation grows. In America, this can be seen in the programs for wiring up of schools, hospitals, and libraries. The programs for the first two already are budgeted at 2.65 bil per year on the federal level, in addition to the various state and local contributions. Their expansion into additional programs is likely, as for example the connectivity for school children from poor families, and for old and sick people at home. Thus, one should expect an expansion of the universal service connectivity programs.

Pricing

Pricing will become increasingly distance-insensitive and usage-insensitive. Pricing will be substantially flat, based on some reserved capacity. On top of that, congestion pricing will deal with peak loads. Congestion is likely to lead, on the architectural level, to a redesign of the network and its switches to divert traffic from the LEC switches, and toload re-balancing. On the economic level, congestion charges will be instituted.

Jurisdiction

The Internet is geographically indeterminate, and hence the question is who, if anybody, will regulate it. This has implications for traditional PTOs which have built up over the years relationships with their national regulatory system, and which now must deal with new agencies. Initially, the regulatory authority of traditional telecom regulatory agencies will be unclear. However, it is likely to be obtained by these agencies to continue traditional mandates. An expansion to new mandates and goals, however, might be blocked by law and politics. Within that realm, there will be a jockeying for the regulatory powers. In America, between Washington and the state. In Europe, between the national NRAs and Brussels. And internationally between various super-national bodies such as the ITU and the WTO. In most cases, regular legal rules will apply to the use of the Internet, such as rules against fraud or securities violations. Some of the Internet regulation will move to intergovernmental bodies. But much of it will shift to self-regulatory bodies of uncertain legal authority in which PTOs will be surrounded by many other parties experienced in fast-paced Internet cooperation. This is exemplified with the debate over the system of domain names. One role for the state will be to assure that the self-regulation will not become a form of cartel management. All this means that the PTOs will be involved on many more levels of rule making than before.

Interconnection

As the system of multiple networks and service providers gains in complexity, it becomes critically dependent on rules of interconnection. The rules are utilized especially against the traditional PTO because its prevalence gives it a special market position and potentially a bottleneck role. It is also the battleground on which the new entrants seek preferable treatment, while the incumbents seek to protect their traditional dominance and cost structure. The specifics for interconnection will involve pricing; unbundling; numbering; and many more. On the whole, the outcome of the rules will depend on the policy priorities of the regulator: where its goal is to introduce competition rapidly, the rules will be biased toward the entrants. But where the goal is to protect the traditional (unreformed) structure of universal service and employment, the incumbent will be favored.

The term "interconnection" covers a wide matrix of relations. On the physical level, that of connective transmission conduits, they include linkages within and among various types of entities:

- traditional and new local telephone companies

- traditional and new long‑distance carriers

- wireless carriers

- domestic and international carriers

- private networks of organizational and user groups

- computer local area and wide area networks

- telephone, computer, and video equipment

- cable television, broadcast, and telecom networks

On the higher levels of applications and content, interconnection becomes an issue of access and interoperability of entities such as:

- Internet service providers

- enhanced (value‑added) service providers

- data and information providers

- video program channels

Unbundling

Where does interconnection take the network system? Interconnection is fairly meaningless without reference to where such interconnection would take place physically. If an incumbent network offers an entrant interconnection at a far-off point, little is resolved. New entrants and service providers want interconnection and access at many intermediate points of the network, i.e., to have it unbundled.

For these reasons, the regulation of interconnection and unbundling go hand-in-hand, before the logical results of competition.

The physical unbundling is only one step in a more complex evolution. The related step will be software unbundling, in which software by outsiders could be put into the central exchange, or interoperate with it from a distance.

Media Regulation

Since the Internet is becoming a mass entertainment medium, questions emerge as to its regulatory treatment: like broadcast TV? Cable TV? Print publishing? Media regulation, even more than telecom regulation, has been politically sensitive, because it directly affects culture and politics. Content issues abound: the protection of children from sexually explicit materials, or from violence, by blockage or a labelling system; national culture and the prevalence of foreign import; accuracy and fairness; privacy protection. All these issues will be dealt with by governments, legislators, regulators, and courts.

Why the Internet will be Regulated

As electronic transactions over the Internet become important the question arises whether they will be controlled on the national and international levels. Many Internet enthusiasts dismiss this question as irrelevant. They believe in the myth that “you cannot regulate the Internet.”

The fallacy is to believe that the Internet is only electronic, which indeed is hard to control. But communications are not just a matter of signals but of people and institutions. “Virtuality” is an appealing notion. But one should not forget that physical reality is alive and well. Senders, recipients, and intermediaries are living, breathing people, who live somewhere in real space, or they are legally organized institutions with physical domiciles and physical hardware. The arm of the law can reach them. It may be possible to evade such law, but the same is true when it comes to tax regulations. Just because a law cannot fully stop an activity does not prove that such law is ineffective or undesirable.

The Internet is not the only priority in any country. The United States, for example, has long worried about crime, Communism and children. That has led it to seek regulatory rules on pornography and encryption, and the wiring of schools for the Internet championed by Vice President Al Gore.

Similarly, Bavaria cares about hate speech and public morality, and sets rules accordingly. For the French, language and culture are priorities, and Internet services originating in the country must be in French. Singapore worries about order. And so on.

For all the rhetoric of an Internet “free trade zone,” no country will readily accept an Internet that includes Thai child pornography, Albanian tele-doctors, Cayman Island tax dodges, Monaco gambling, Nigerian blue sky stock schemes or Hong Kong junk E-mail purveyors.

Thus, for better or worse, each society will apply its accumulated wisdom, misconceptions, preferences and interest group muscle to the rules governing transactions over the Internet. And these rules will not be very different from those applied to the rest of society.

Columbia Institute for Tele-Information

May 1998

Presented at the

Introduction