Home - Key Questions - Fixed Networks: Section 2 - Question 9

Section 2: Questions regarding the drivers of change in the network layer, Layer II

Question 9: Does the economics of the network layer imply that this layer will be dominated by a handful of large global operators?

There are two stories that may be told in answer to this question. According to the first story, there are several economic forces driving the network layer that will produce an outcome where only a handful of large global network operators will come to dominate this part of the new telecoms industry.

The second story argues that while these economic forces might exist they are not likely to be as powerful as this first story suggests. Moreover, other economic forces will co-exist that will make it easier for a larger number of firms to continue to contest markets in the network layer. Accordingly, rather than this layer being dominated globally by only a handful of firms, we will see many firms competing in these markets, including small and medium sized firms concentrating on niche areas. What are these economic forces?

The First Story

In the first story emphasis is put on the nature of the costs that are incurred in constructing and operating telecoms networks. More specifically, it is argued that a significant proportion of these are fixed costs, sunk costs and shared costs, and usually substantial excess capacity is built into these networks.

Fixed costs are costs that do not change as output - in this case telecoms services - is increased. For example, an operator will have to commit itself to substantial costs in constructing its network and developing management, operating, marketing and sales capabilities. Over a certain range of output these costs will be fixed irrespective of the number of customers the operator has and the traffic generated. For instance, the costs of providing the initial switches and transmissions systems will be the same regardless of whether the network is providing services for only one customer with little traffic, leaving a great deal of surplus capacity, or whether it is servicing many customers and operating at full capacity.

Sunk costs are costs that cannot be recovered once a company ceases operating. For example, some of the equipment or software may be specifically tailored to the needs of the particular operator and may not have alternative uses.

Shared costs are costs that are split between two or more services. For instance, it is often possible to provide both voice and data services over the same network. Furthermore, networks are often constructed with significant excess capacity. For example, Qwest, in laying its optical fibre network alongside rail-tracks in the US, made the decision to lay two conduits, each of which is capable of carrying several optical fibre cables containing around 96 optical fibres each. The economic logic behind this decision is that once the cost of digging a cable trench has been incurred it costs very little extra to lay a second conduit, significantly less than it would cost at a later stage to dig another trench in order to expand capacity. In 1998 it was estimated that Qwest was utilising only around one per cent of the total capacity of its network despite having expanded its customer base and revenue substantially.

High fixed and sunk costs together with excess capacity imply that an operator’s average cost (i.e. unit cost) will fall as it acquires more customers and carries more traffic over its network. In other words, the operator will enjoy the benefits of increasing returns and economies of scale. All else being equal, an operator that can increase its customers and traffic more than its competitors will be able to operate with lower costs and higher profitability than its rivals. Accordingly, this operator will be in a better position to expand in the next period at the expense of its less efficient competitors. Furthermore, an operator who can spread the total network costs (including marketing, sales and other expenses) over multiple services may be able to provide these services at lower cost than a rival selling only one service, thus benefiting from economies of scope⁽¹⁾ .

According to the first story, with increasing returns and economies of scale and scope, a few successful firms - whatever the reasons for their initial success - will attract more customers and traffic than their rivals and in doing so will drive their costs below those of their competitors. This will make them even more competitive relative to their peers which will allow them to expand their output (services) even more, leading to a further reduction in relative unit costs, etc. In turn, this will lead to a process of cumulative causation that will ultimately result in this part of the industry being dominated by a handful of firms.

The Second Story

The second story paints a very different picture of the evolution of the network layer. This does not deny the existence of increasing returns, economies of scale and economies of scope. But it argues that the extent of these factors is less than that assumed in the first story with the result that the impact of these economic forces will not be as great as this story suggests. This is so, it is argued, for several reasons.

First, the extent of fixed costs is not nearly as important as suggested. For example, a new operator does not have to construct its own end-to-end network, connecting senders and receivers of calls, before it can begin operating and competing. Indeed, it does not even need a complete network. As a result of government regulations that provide for the purchase and resale of network capacity and for interconnection to the networks of other operators, it is possible for an operator to enter the industry incrementally. That is the operator can begin by installing a few switches that will allow it to route its traffic in the way it wants while depending on the networks of others in order to originate and complete its calls. The more excess capacity there is and the stronger the degree of competition between network operators, the closer will be the price of access to their networks to the marginal cost of these operators.This will mean that a new operator will not necessarily be at a significant cost disadvantage compared to other companies that have started business earlier and are operating on a substantially larger scale.

As a new operator learns from its experience, becomes more efficient and successfully establishes its brand name, it may then be able to incrementally expand its own network, financed from its revenues, thus decreasing its dependence on the network services of others. This is precisely the way in which all new operators have entered the new telecoms industry. According to the second story the fact that they have been able to enter, despite the relatively small initial scale of their operations, is proof that increasing returns and economies of scale and scope have not given overwhelming competitive advantages to larger scale operators. Or at least this appears to have been the case while new operators have continued to successfully enter the market.

The second argument put forward in the second story is that the importance of new technology has been seriously underestimated in the first story. While successful larger scale operators may be able to ‘charge down the falling long run average cost curve’ (i.e. benefit from falling unit costs as their output increases), they may at the same time be locked into older systems that are being outperformed by more recent generations of technology. Encumbered by legacy systems (i.e. networks that depend on older technologies) they may not be able to take full advantage of the latest technologies. Therefore, older larger scale operators will not enjoy the extent of benefits suggested in the first story. If new technologies are able in this way to periodically refresh the entry process, the advantages to larger scale firms of increasing returns, and economies of scale and scope, will be reduced. Accordingly, rather than a handful of firms dominating Layer II of the industry there may be many more firms, each benefiting to different degrees from the latest technologies.

A concrete case where such new technologies have substantially benefited new operators at the expense of incumbents has been self-healing ring SDH/SONET networks. In the UK, for example, operators like COLT (owned by the US mutual fund company, Fidelity) and Energis (owned by the National Power Grid) were able, entering in 1992 and 1993 respectively, to build their networks entirely on the basis of this technology. On the other hand, the incumbent, BT, was partially constrained by its PDH/SDH hybrid network (PDH being the older technology). This technological edge gave the new operators significant advantages such as lower cost, greater reliability, flexibility in terms of bandwidth provisioning and quicker repair times. This may also explain some of the performance advantages enjoyed by these and other new operators vis à vis BT as recorded in audited performance measures made publicly available by the UK regulator, Oftel (see the Comparable Performance Indicators web site for further details).

But this was not the end of the story. Influential financial analysts, such as Salomon Smith Barney, seized on these competitive advantages in their stock market valuations of companies like COLT and Energis in order to argue that they were likely to enable the companies to attract customers away from BT and win significant shares of the UK market. In turn, such reports contributed to the astronomical rise in their share prices. Indeed, in 1998 and 1999 COLT and Energis were amongst the best-performing shares on the London Stock Exchange.

The next argument in the second story is that the emphasis solely on costs and price competition in the first story is misleading. While these are an important determinant of competitiveness, there are other determinants that may be equally, or in some cases even more, important. Examples include reliability, security, flexibility in bandwidth provisioning - in other words the performance characteristics that can be provided by new technologies. However, these are not the only determinants of competitiveness apart from cost and price.

Further determinants include customer care and customer responsiveness. Knowing that your telecoms operator will not simply connect you to its pipes and carry your data, but also understands your information and communications needs and is committed to working with you to develop solutions to your problems in these areas may be one of the most important determinants of competitiveness in the business market. However, the ability to provide superior quality of service and customer care and, most particularly, the ability to understand the customers' needs and problems and devise appropriate solutions may bear little relationship to scale of output. A company that is relatively small, but highly focused on particular market segments, may therefore be more competitive and perform better than one simply relying on falling costs due to increasing returns and economies of scale and scope.

The final argument in the second story is that while there may be advantages of scale, there may also be disadvantages. For instance, while the major incumbents benefit from their scale of operations, their financial strength, their brand name, their existing customer base, and the other advantages that incumbency provides (such as the opportunity to drag their feet in implementing regulatory measures), they may also suffer from disadvantages of size. These might include, for example, hierarchical, bureaucratic and inflexible corporate structures and a lack of focus as a result of covering most areas of their domestic markets. On the other hand, smaller new operators may enjoy the benefits of 'F-4 Firms': being focused, flat, flexible and fast. Once again the implication is that, contrary to the first story, it is by no means inevitable that the network layer of the new telecoms industry will be dominated by a handful of large global operators.

What do you think about these two opposing stories? Which do you think will provide a better prediction about the future of the industry over the next five years?

If you wish to express your views on questions such as these go to the Workshop (Area 1). To compare your visions with those of others go to Vision Check.

(1) Economies of scope exist when a firm is able to sell two or more products or services more cheaply together than if they were sold separately. It is economies of scope that lie behind the preference of most large telecoms companies to sell 'packages' of services such as local, long distance and international calls, mobile and internet access.