The Economics of Networks

1. The Economics of Networks Nicholas Economides

2. Abstract • Paper Purposed to analyze the salient features of networks and similarities between the economic structure of networks and structure of vertically related industries. • Analysis focuses on positive consumption and production externalities, commonly called network externalities. • Analyzed compatibility, coordination to technical standards, interconnection and interoperability, and their effects on pricing and quality of services and on the value of network links.

3. Motivation: Why Bother? • Importance / Significance of Network Industries: • Modern Economy would be diminished without transportation, communications, information and railroad networks. • Non-network industries’ characteristic’s are closely related to Network Industries. • Therefore we can apply lessons of networks to industries where vertical relations play a crucial role.

4. Network Classifications Networks are formally composed of links with connect nodes. From the diagram to the right, we can see that many components are needed for providing a good or service. Therefore components of a network are complementary of one another. There are usually close substitutes for components in networks. For example, transmission can be done through a cable TV line, fixed telephone line, wireless satellite, OCN, etc.

5. One- Way or Two- Way Networking: Telephone network example This type of network is not a function of the topological structure of the network, but depends on the interpretation of the structure to represent a specific good or service. The Figure to the right is interpreted as a two way telephone network where SA represents a “local switch” in city A, Airepresents customer in city A, and similarly for SB and Bj. .A long distance phone call would be conveyed by AiSASBBj. .We can also interpret this as an ATM transaction from Bank Bj from ATM Ai, where is it conveyed as AiSASBBj. Connections AiSAAk and BjSbBi may be feasible, but there is no demand for them.

6. Network Externalities • Networks exhibit positive consumption and production externalities. • A positive consumption externality signifies the fact that the value of a unit of the good increases with the number of units sold. • To economists, this seems counterintuitive, since we know market demand slopes downward- with the exception of Dr. Rams’ lectures, of course. (mumble mumblemumblemmmmhmmhmmm)  • Therefore, we should thus interpret positive externalities effects as “the value of a unit of a good increases with the expected number of goods sold.”

7. Sources of Network Externalities Depending on the network, the externality may be direct or indirect. When customers are identified with components, the externality is direct. In the figure to the right, we have an n- component network, where there are n(n-1) potential goods. An additional (n +ith) customer provides direct externalities to all other customers in the network by adding 2n potential new goods through a provision of a complementary link (say ES) to the existing links.

8. The ‘Macro’ vs. ‘Micro’ Approach Approach • Macro Approach presumes that network externalities exist and attempts to model their consequences. • This approach is easier and produces strong results. • Micro Approach attempts to find the root cause of network externalities. • This approach is more difficult, and more constrained in many ways, however hosts a significant benefit in defining market structure.

9. Perfect Competition: Macro Approach • Externalities arise naturally in both one- and two-way networks, as well as in vertically related markets. • The value of good X increases as more of the complementary good Y is sold, and vice versa. Thus more of Y is sold as more X is sold. • Why is this idea not explosive? • Positive feedback loop is not explosive because of the inherent downward sloping demand curve (the consideration of consumption expectations).

10. Graphically Understanding the downward sloping demand curve in the light of externalities. If we let the willingness to pay for the nth good when ne units are expected to be sold be p(n;ne). This is a decreasing function of its first argument because the demand slopes downward. P(n;ne ) increases in ne ; This captures the network externalities effect. We should note that as our ni increases, we consume on a different demand curve. Due to Diminishing marginal returns, our expectations of how much we are to consume decreases, but we will continually go up to higher demand curves due to the persistence of positive externalities. I.E. The option to consume at greater combinations will be available, but it will not be chosen. We should note that there is a MUCH larger external benefit to society than to individual firms here due to DMR.

11. Some Enlightenment • Therefore, in the presence of network externalities perfect competition is inefficient. • Marginal social benefit of network expansion is larger than the benefit that accrues to a particular firm under perfect competition. • THUS, Perfect competition will provide a smaller network than what is socially optimum, and for some relatively high marginal costs perfect competition will not provide the good while it is socially optimum to provide it.

12. Monopoly- No, not the game • Economides and Himmelberg (1995) show that a monopolist who is unable to price discriminate will support a smaller network and charge higher prices than perfectly competitive firms. • Influence over expectations drives monopolist to higher production, but the monopolist will tend to profit maximize towards inefficient levels of production that are lower to that of perfect competition.

13. Oligopoly and Monopolistic competition under compatibility This Analysis can easily be extended to monopolistic competition among compatible oligopolist’s if firms face downward sloping average cost curves. Firms produce on the downward sloping part of the firm scaled fulfilled expectations demand. At equilibrium, firm j’soutput is determined at the intersection of MC and MR. At monopolistically competitive equilibrium, the AC curve is tangent to the fulfilled expectations demand at qj.

14. The ‘Micro’ Approach : Analysis of specific micro-structure of a network We identify cases where Only end-to-end services are demanded and the case when there is also demand for some services that do not reach from end-to-end. To the right is a simple case where only end-to-end services are demanded. Goods A and B are complementary. Each type of good has a number of brands available, Ai, i=1,…,m, Bj, j=1,…,n. Complementarity is actualized when the components Ai and Bj are combinable and function together without extra costs; i.e. when they are compatiable.

15. Compatibility VS Incompatibility Economides analyzes the choice of asset ownership as a consequence of the choice of technology (and the implied degree of compatibility). This creates a three stage game of compatibility choice in the first stage, vertical integration 9or disintegration) in the second, and price choice in the final stage. In analyzing the stage of compatibility choice, the influence of the anticipation of decisions on (vertical) industry structure on compatibility decisions are evaluated (refer to diagram to the right).

16. Compatibility VS Incompatibility Each B-Type component is immediately compatible with either A1 or A2. Each brand Bi produces two version, one compatible with A1 and one with A2.

17. Network Externalities and Industry Structure • When there are strong externalities present, the monopolist exclusive holder of a technology may have an incentive to invite competitors and subsidize them. • They may be unable to credibly commit to high output as long as he is operating by himself. • Since the level of market depends mainly upon other firms, the commitment to high output is credible.

18. Network Externalities: Invitations to Enter • The invitation to enter and the resulting increase in market output causes a competitive effect and a network effect. • Competitive Effect • An expected increase in competition due to increase in number of firms. • Network Effect • Tends to increase willingness to pay and market price because of higher expected sales.

19. Concluding Remarks: Is it 1:45 yet? :-/ • Nicholas Economides noted some interesting issues that arise in networks and vertically related industries (especially in the presence of a fragmented structure). Many open questions are still prevelent, nonetheless. • One Large issue still unresolved is the joint determination of an equilibrium market structure coupled with the degree of compatibility across firms.

20. Conclusion • On a more fundamental level, there is no good prediction yet on the ‘break points’ that define complementary components in a modular design structure. Even if these points are known, little analysis has been done of competition in a multilayered structure of vertically related components. • Nevertheless, it is precisely this kind of modeling that is needed for an analysis and evaluation of the potential structures of the internet.