ساختار بازار، مقررات و سرعت نفوذ شبکه تلفن همراه

The speed of market penetration (i.e. diffusion) is an important summary measure of how well the market works for potential consumers of a new product. This paper identifies the structural features associated with rapid diffusion of mobile telephony. We use a sample of thirty countries over the 16 years in which average penetration rose from 2% to 97% of the population (earlier studies observed only the initial years of diffusion during which there was typically only one or two networks). We find that both the number of networks and the history of market structures matter for the speed of consumer uptake. The market structure effect does not appear to work exclusively through the level of prices. Digital technology, standardization, privatization and independent regulation are also important positive factors, and we identify the speed and dimensions of catch-up.

In the context of a growing market with consumer network externalities, the speed of a new product's market penetration (i.e. diffusion) is an important summary measure of how well the market is performing for potential consumers. Delays in uptake can result in large welfare losses.1 When the market is regulated, it is particularly important to understand how the various potential regulatory levers (e.g. number of firms, public ownership, price controls) affect the diffusion process. As penetration approaches saturation, usage becomes the more important indicator of market performance, but new consumer products continue to be developed and their speed of diffusion is of considerable significance. In this paper we focus on understanding the central period of diffusion in which average market penetration across the more developed economies rose from less than 2% to nearly 97% over 16 years. Like any other product, the demand for mobile phone services is influenced by a range of marketing and technical factors that constitute the overall product ‘offer’. This offer includes price level, price structure (e.g. cost of sending relative to receiving a call), reach (geographic coverage) and reliability. Individual elements of the product offer are difficult to observe and measure on a consistent basis either internationally or over time. Furthermore, the optimal balance in the offer can be sensitive to national idiosyncrasies. In fact, one of the theoretical virtues of a competitive market is that it creates incentives for firms to respond to these idiosyncrasies and to provide the most attractive offer to consumers. This leads us to focus on the structural characteristics of the market that drive competition. The aim of this paper is to identify those structural features that are associated with the competitive environment which maximizes the rate of diffusion of mobile telephony through the population. Mobile network penetration has been expanding rapidly in recent years, though there are signs it is reaching maturity in the advanced countries. We employ a panel of 29 OECD countries and China over the period 1991–2006. We include China because of its scale and economic growth at the time, but we also test for robustness using the OECD-only sample.2 This period covers the core of the penetration phase in each market.3 We focus on three structural features: the number of firms; ownership (i.e. privatization); and the existence of an independent industry regulator. Although the number of mobile networks is tightly regulated, we also investigate the potential endogeneity of market structure. Earlier work on telecom market penetration (including fixed line) focused on demographic and technology factors, privatization, first new entry and the early part of the diffusion process.4 The latter two limitations appear to be important because, by using data that more completely covers the core diffusion years for the countries in our sample and distinguishing between different numbers of firms, we find substantial effects beyond the simple monopoly versus duopoly dichotomy.5 Thus, while previous work has typically found that opening the market beyond monopoly is beneficial, it provides little guidance for important competition policy issues such as the number of operators to be licensed or merger regulation. The previous empirical literature also has little to say about regulatory institutions. Our main contributions are to distinguish the fine-grained effects of each extra entrant and technological factors, and to estimate our model over the core years of the diffusion process.6 Having identified the key structural features associated with rapid diffusion, we go on to ask whether the effect of a more competitive structure works mainly through the average price level as distinct from non-price-level elements in the offer. The remainder of this paper is structured as follows. In the next section, we review some related literature on competition, ownership and regulation in telecoms markets. Section 3 sets out the econometric methodology and Section 4 describes the data. Section 5 presents and discusses the empirical results, including tests for endogenous market structure. Section 6 concludes.

The aim of this paper is to identify the structural features of a partly regulated market that provide the best competitive environment to maximise the market penetration of a new product – mobile telephony. Unlike earlier studies, we are able to use data that covers the core period of the diffusion process. Our specification does not impose a functional form on the effects of alternative numbers of mobile networks. Like earlier studies, we confirm the benefit of moving from monopoly to duopoly, but the advantage of using more recent data is that we now have experience of a much wider range of market structures. This reveals that pentopoly (i.e. five firms) is a major competitive improvement on duopoly. There is no evidence for further improvement in diffusion with more firms. More provisionally, we find that market structure effects do not appear to operate exclusively through prices – there are other elements of the product offer, also related to market structure, that affect consumer uptake. It is interesting to relate this to the wider empirical literature that relates market structure and competitive outcomes. Much of this now exploits data on local geographical markets to investigate the implied effects of competition on margins, price and productivity.44 An emerging stylised fact is that a relatively small number of firms, often between two and four depending on the product, is sufficient to generate most of the benefits of competition in traditional homogeneous product markets (e.g. professional services, retailing, concrete). That research uses genuinely local markets (compared with national markets in this paper) to provide the cross-section dimension. It also investigates totally different dependent variables. Nevertheless, our results are consistent with the view that relatively few firms may be sufficient for competition in relatively homogeneous product markets. In our case, that number is five. There is an important difference between mobile networks and more traditional markets because spectrum limitations have been addressed by strict licensing of operators. This eliminates the threat of entry as a mechanism by which competition works. The institutional response has been that mobile networks are typically regulated, though the independence of the regulator has varied across countries and over time. We find that the independence of the regulator has a modest positive role to play in addition to market structure. In line with some of the earlier literature, privatization also has a substantial positive impact on diffusion. Our findings are consistent with the view that a balance may need to be struck between investment incentives for network industries characterised by large sunk costs and the benefits of an apparently more competitive market structure, but this balance may require five firms. This is particularly relevant when determining the number of spectrum licenses to be granted, but it is also relevant for merger policy. Our findings additionally support the view that private ownership and independent regulation are desirable in the absence of an entry threat. Our results also cast light on how consumers respond to competition between multiple standards. We find that diffusion is faster when there is standardisation. This may be due to either customer confusion or rational delay in adoption until a dominant network technology emerges. Finally, the data in this paper covers the core period of diffusion in thirty countries. Average market penetration across these countries rose from less than 2% to nearly 97% in 16 years. As the mobile network market matures, the consumer focus naturally turns to usage and product development.45 It is not obvious that competition at the mature stages of the product life cycle will be determined in the same way as at the diffusion stage. However, new products are always emerging and it remains important to have developed a more complete view of the role of competition in the diffusion process of a new consumer product.