سرمایه گذاری های راهبردی با سرریزها، ادغام عمودی و حق اقامه دعوی در بازار دسترسی به پهنای باند

We analyse competition between two retailers of broadband access when they differ in their ability to offer value-added services. One retailer is vertically integrated and controls the input-market for local access. This firm invests to increase the input quality (upgrading to broadband) before an access price regulation is set. We first show that access price regulation may lower consumer surplus and welfare if retailers do not differ too much. Second, if the integrated firm’s ability to offer value-added services is much higher than that of the rival, the integrated firm uses overinvestment as an alternative foreclosure tool.

The purpose of this paper is to examine the interplay between a facility-based vertically integrated firm and an independent competitor in the retail market for broadband Internet connectivity. The latter firm buys local access as an input from the former firm. The vertically integrated firm undertakes an investment (broadband upgrades) that increases the quality of the input. We assume that the regulator has only one instrument available, an access price regulation for the input sold to the independent rival.1 The retail market is assumed to be unregulated.2 Furthermore, we assume that the access price is set after the investment but prior to retail market competition since the regulator has limited commitment ability. Both the timing structure and the one-sided regulation of the input segment correspond to the dominant regulatory paradigm in the EU and the USA (Laffont and Tirole, 2000, Hausman, 1997 and Cave and Prosperetti, 2001). Installation of fiber in the local access network will be a substantial, lumpy, and irreversible investment, and the economic life of the investment will be longer than the regulation contract used for access prices (Hausman, 1997).3 The access price regulation may reduce investment incentives, and the main message of this paper is that the total welfare effect of access price regulation critically depends on which firm has the highest ability to transform input to output. The quality of the input component sold from the integrated firm is the same for both retailers, but the retailers may differ in their ability to offer value-added services (broadband services such as interactive video).4 Except for the case where the independent firm has the highest ability to use the improved input quality, the integrated firm will foreclose the rival from the market through the access price in an unregulated environment. However, this is not a sufficient condition to ensure that an access price regulation improves consumer surplus and total welfare. If the retailers do not differ too much with respect to their ability to offer value-added services when the input quality is improved, we show that access price regulation reduces the vertically integrated firm’s investment incentives. An access price regulation lowers consumer surplus and total welfare as long as the cost of investment is not too convex. If the vertically integrated firm’s ability to offer value-added services is much higher than that of the independent rival, an increase in the investment will reduce the quantity offered by the independent retailer. An access price regulation still eliminates the vertically integrated firm’s ability to use the access price as a foreclosure tool, but now the integrated firm may use overinvestment as an alternative tool to drive the rival out of the market. Today the majority of residential consumers use their telephone lines for the last mile of narrowband Internet connectivity, and by upgrading their local networks the telecommunication providers are able to increase the speed of communication.5 The high up-front investments of new wire line facilities, and the possibility of increasing the capacity and quality of existing local telephony and cable-tv networks, indicate that telephone companies and the cable-tv-companies that already have installed wires to homes, will control the segment for broadband local access to residential consumers (Mackie-Mason, 1999).6 In the current regulation only the telephone access provider is mandated to supply local access as an input to non-facility-based rivals in the retail market. Therefore, the telephony incumbent has been the only provider of local access as an input to independent ISPs7. Hence, given the existing asymmetry in regulation of the telecommunication and the cable-tv technologies (see Hausman et al., 2001), the present model only fits for services offered by the telecommunication incumbents.8 We assume that the investment in higher speed of communication may be seen as an unambiguous improvement of quality.9 The closer to homes the fiber is installed, the higher is the quality. The trade-off between the distance the existing lines are used and the network quality (speed) implies that the upgrading costs are convex in speed of communication. In Fig. 1 we illustrate the Internet as a layered network with the physical network as the bottom layer. Local access is obviously an essential input component for the ISPs. The functions of the retail ISPs are to combine the components’ local access, regional backbone capacity and global backbone capacity, and they act as a kind of portal to the applications and content in the Internet. Full-size image (4 K) Fig. 1. The layered structure of the Internet. Figure options As described by Cave and Mason (2001) and Faulhaber and Hogendorn (2000), the retail ISPs must choose their regional and global backbone capacity before they serve the end users. This implies capacity constraints that limit the number of consumers that can be served.10 With respect to timing in our model it is reasonable to assume that the investment choice of speed of communication in the access network is taken prior to the ISPs’ choice of local and global backbone capacity. The interplay between local retail ISPs and the upstream providers of global access will not be addressed in the present paper. According to the Chicago School there is only one profit to be extracted when two goods are complementary. Therefore, when an upstream monopolist can exploit its market power in the upstream market, a vertically integrated firm has no incentives to foreclose retail rivals. However, in the present context, the vertically integrated firm foreclosure incentives are driven by the inability to exploit the monopoly power in the upstream segment. First, even without access price regulation, we assume that the integrated firm sets a linear input price. Hence, there will be a double marginalization problem due to imperfect competition in the downstream market. Second, we analyse the case where the monopoly input is price regulated, while the downstream market is unregulated. A vertically integrated firm which is forced to sell the input equal to or close to marginal costs has an incentive to use non-price methods to deny rivals access in the unregulated downstream market. Rey and Tirole (1997) analyse the incentives for foreclosure by a vertically integrated firm that controls an input bottleneck in an unregulated market, while Laffont and Tirole (2000) discuss the incentives for non-price discrimination under access price regulation. In a context where a vertically integrated upstream monopolist faces an access price regulation, several recent papers analyse the use of non-price foreclosure towards downstream rivals. Analogous to the present paper, Economides (1998) assumes an exogeneously given market structure where an integrated firm controls the input-segment, and where there is an unregulated Cournot duopoly in the retail segment. Economides (1998) shows that with an exogenously given access price the integrated firm will always use non-price foreclosure towards the retail rival.11 This result contrasts with Sibley and Weisman (1998) and Weisman and Kang (2001), who find that the vertically integrated firm will be less inclined to degrade the quality of the input if the profit margin is high in the input segment. All these papers assume Cournot competition in the retail market, and that the foreclosure activity is assumed to degrade the quality of the input sold to the rivals. In contrast, in our model there is no opportunity to unilaterally reduce the quality of the input sold to the rival. The quality level of the input is the same for both retailers. However, if the vertically integrated firm has significantly higher ability to offer value-added services than the rival, the integrated firm may commit itself to be more aggressive in the retail market by overinvesting in network quality improvement. Beard et al. (2001), Weisman, 1995 and Weisman, 1998 and Reiffen (1998) analyse the incentives to use quality degradation if there are Bertand retail competition and horizontal differentiation.12 The outcome compared to Cournot competition is ambiguous. Bertrand implies more intense retail competition, and as long as the rival is not completely foreclosed from the market, this will reduce the incentives to use quality degradation. When the degree of horizontal differentiation increases, the integrated firm cannot replace the rivals’ sale with its own retail sale, and both price and non-price methods to induce foreclosure will be less profitable.13 Faulhaber and Hogendorn (2000) and Foros and Kind (2003) analyse the broadband access market with focus on the choice of target market (where to upgrade to broadband), while the coverage decision is not analysed in the present paper. Another distinction from the present paper is that these two papers develop models of competition among several facility-based broadband access providers, while we analyse the interplay between a facility-based provider and a non-facility-based rival. Hausman et al. (2001) analyse the asymmetric regulation of telecommunication providers and cable-tv providers regarding broadband access. Rubinfeld and Singer (2001) analyse the merged AOL Time Warner’s incentive to engage into two types of non-price foreclosure in the broadband access market. The article is organised as follows. In Section 2 we present the model. In Section 3 we give some concluding remarks.

In this paper we have analysed a market structure where a vertically integrated firm controls an essential input for retail providers of Internet connectivity. The vertically integrated firm may undertake an investment that increases the quality of the input (upgrading to broadband). In an unregulated retail market the vertically integrated firm competes with an independent firm that buys access as an input. We analyse the effect of an access price regulation that is imposed on the vertically integrated firm. Since an upgrade of the local access network to broadband is an irreversible investment, we assume that the regulator has limited commitment ability with respect to the access price. Hence, the access price is set after the investment, but before retail competition. We compare the access price regulation regime with the outcome without regulation. The total effect on consumer surplus and welfare critically depends on whether the vertically integrated firm has higher ability to offer value-added services (broadband services) than the rival firm. If the retailers do not differ too much with respect to their ability to offer value-added services, when the input is improved, we show that access price regulation reduces the vertically integrated firm’s investment incentives. Furthermore, an access price regulation lowers consumer surplus and total welfare as long as the cost of investment is not too convex. In contrast, if the vertically integrated firm’s ability to offer value-added services is much higher than that of the independent rival, an increase in the investment level will reduce the quantity offered by the independent retailer. The vertically integrated firm may then use overinvestment as an alternative tool to drive the rival out of the market. In our model the regulated access price will be set equal to or close to the marginal cost. Hence, compared to the case without access price regulation the regulator removes most of the vertically integrated firm’s cost advantage in the retail market. If the retailers’ ability to offer value-added services is quite similar, then the independent firm will have higher profit than the facility-based firm since the latter has to cover the investment costs. Put differently, the access price regulation may imply a second-mover advantage. In the present paper we have not focused on entry, but this feature of the regulation will probably discourage facility-based entry. In particular this will be true if we incorporate uncertainty in the analysis. The non-facility-based firm may enter the market later, and, furthermore, does not need to make the irreversible investment. The timing structure seems to correspond to the current regulatory paradigm both in the EU and the USA. This paradigm mandates that the access price should be set to the long-run incremental costs (LRIC). At first glance, this may include the investments in, e.g. broadband upgrades, while we show that the regulator will set the access price equal to the marginal cost as long as the retailers do not differ too much in their ability to offer value-added services. The main feature is, however, that the determination of the long-run incremental costs is highly discretionary and that the decision is taken after the investment is made. Hence, its impact on incentives to invest before the discretionary decision on the access price will be analogous to the case analysed in the present paper. Hausman (1997) argues that FCC’s measure of LRIC ignores the existence of technological progress that is reducing the prices and increasing the quality of the components in a broadband access network. Hence, LRIC implies an access price corresponding to the most efficient components available at the time the access price is set, and this will not cover the investment costs under a rapidly changing technology. Uncertainty is not formerly analysed in the present model, but there is obviously a significant probability of failure when a firm invests in a broadband access network, and this will make the problem of investment incentives even more important. The current LRIC approach does not cover the costs of unsuccessful services.23Cave and Prosperetti (2001) focus on the situation in the EU, and they argue that the incumbents do not have sufficient incentives to upgrade to broadband access, since they anticipate that they will be forced to offer access at cost-based prices. While the LRIC-approach may be detrimental to the vertically integrated firm’s investment incentives, the vertically integrated firm’s retailer may have a competitive advantage from the way the linear access price is designed. The forward looking incremental cost is incorporated, and the rival is then offered an access price equal to an average cost that is above the marginal costs. Hence, with unregulated Cournot competition in the retail market, the vertically integrated firm’s retailer bases its reaction curves on the marginal cost, while the rival needs to pay the average cost. Hence, in presence of regulation, the vertically integrated firm may prefer that the input is charged with a linear price rather than with a non-linear access price that consists of a fixed fee combined with a unit price equal to (or below) marginal costs.24 If there was no ability to improve the network quality through the investment, a regulated access price close to or equal to marginal costs would improve welfare compared to the case without an access price regulation. This indicates that the current regulation regime in the EU and the USA may be better than no regulation if static efficiency was the only goal for the regulator. In contrast, when there is an investment decision before the access price regulation is decided, the benchmark without access price regulation may imply higher welfare, but the conclusion is ambiguous. Obviously, the regulator may alter the outcome by non-linear wholesale prices, price regulation in the retail market, and line-of-business restrictions, but the basic challenge seems to be the choice of rules versus discretion in the governments’ policy. When the policymaker has the opportunity to set the access price discretionary after the investment, it will set the access price close to marginal cost (or LRIC). When the decision is taken after the investment, this is the best thing to do for the regulator, given the current situation. It is not a result of non-optimal behaviour from the regulator. If the regulator wishes to realise the outcome in the case without access price regulation, it has to credibly commit to a policy rule before the investment that prevents the regulator from using the discretionary access price regulation.