Competition in Communication Networks: Pricing and Regulation (original) (raw)

2001, Oxford Review of Economic Policy

It is no surprise, then, that the communication sector has attracted the attention of policymakers and economists. But there are further reasons, beyond just the size of the sector, that single out communications as an area of particular interest. These reasons relate to the characteristics of the networks that deliver communication services. The characteristics are not individually unique to communication networks, as the discussion will make clear. The combination of all four factors is, however, unique to networks. We argue in this paper that the combination represents a particular challenge to the study and regulation of communication networks. The first characteristic is the cost structure of networks. Typically, there is a large fixed cost to building a network-digging up pavements to lay cables for a communication network, laying track or asphalt for rail or road travel, gathering all the information for the first edition of an encyclopaedia. For example, the estimated fixed cost of the British Telecom (BT) network in the U.K. accounts for approximately two-thirds of the total cost of the network (Oftel, 1999). Moreover, these costs are often sunk, i.e., irrecoverable, even if production stops. In contrast, the marginal cost of operating the network is low. On a standard telephone network, the marginal cost of making a local telephone call is of the order of 0.2-0.4 pence per minute. The marginal cost of producing an encyclopaedia is the cost of writing and shipping a set of CDs-a couple of pounds. The economies of scale that result from this cost structure often lead to "natural monopoly"-a dominant firm that captures the entire market. There are two observations to make at this point. First, this cost structure occurs in many, non-network industries. On its own, then, this factor does not raise any issues that are not familiar to any economist who has taken an introductory micro course. Secondly, there is often considerable debate about whether fixed costs are all that important in communication networks. In the past, it has been thought that the access part of telephone networks (the last segment of the network, from the last switch to the customer's premises, called the local loop) was such a significant cost that provision of access should be monopolized. Recently, this view has been brought into question. The next section considers this issue in some detail. Suppose that you want to send an e-mail. For most of us, we would first access the Internet by calling an Internet Service Provider (ISP) over a telephone line. The average email message is broken into around 20 pieces, or packets, by the sending computer. These packets are sent over a standard telephone line to the ISP, using a modem to convert the computer's digital information to the analog waves that telephone lines transmit. Each packet is transmitted to the nearest 'router': a special computer, dedicated to receiving and forwarding packets, that is the Internet equivalent of a telephone switch. The router consults a database to decide where to send the packet; it passes each packet onto another router, or to the destination if it is close enough. Once all the packets arrive at their destination, they are reassembled into the original e-mail and read. This seems pretty involved just for a message like "I'll be home in half an hour". Of course, each user does not care about which routers have handled the individual packets in an e-mail. What matters is the joint function of all of the components that go in to sending an email. In the language of economics, the different parts of the network are called complements-items that are worth more together than separately. This is the second factor that distinguishes networks. Again, complementarity is not unique to networks. I do not value particularly the individual components of the starter unit in my car; I care only about whether it starts when I turn the key. Taken with two other factors, however, complementarity is particularly marked in networks. First, positive externalities (about which, see below) mean that there are large gains to connecting two networks. Secondly, the cost structure means that some network segments are owned by a small number of firms, leading to a bottleneck problem. This is the case with the "last mile" of telephone wires running into houses, which typically is owned by an incumbent telephone company. Any telephone company offering services have to gain