140 be offered via broadband access and VoIP technology, the open-access regulation of switching devices will no longer be required.122 8.3.2 The extent of the monopolistic bottleneck in next generation networks There is a controversial debate over the regulation of infrastructure investments that incumbents are currently making into so-called next generation networks (NGNs). The goal of NGN migration is to merge the previously separate voice and data networks into one fully converged network with all IP-based transmission technology. This network will replace the current public switched telephone network (PSTN) and the IP networks. NGN providers plan to offer “triple-play” (voice services, Internet and TV) to their end-users. To this end, NGNs have been defined as networks with bandwidth sufficient to support services that are not supported by current broadband technologies (i.e. simultaneous transmission of high definition television, broadband Internet and voice services). Convergence will decrease network costs as (1) only one network needs to be managed, (2) existing capacity is used more effectively with all-IP transmission, and (3) network maintenance costs are lower in IPbased networks compared to PSTN networks. For the migration to an all-IP NGN network, investments are necessary in local communications infrastructures as well as in long-distance communications infrastructures. In the local network, high-capacity lines need to be built out closer to the end-user. Here the attribute “next generation” generally refers to bandwidth offers of up to 50 Mbps. Generally this is realized by building fiber lines up to the street cabinet (FTTC = fiber-to-the-curb) or up to the end-user premises (FTTH = fiber-to-thehome). In the case of FTTC the bandwidth of the remaining metallic local loop between the street cabinet and the end-user is enhanced with xDSL transmission technology (generally ADSL+ or VDSL). In the long-distance infrastructure of an NGN the existing voice switches need to be replaced by routers and Voice over IP equipment (Marcus, 2006a: 5). The prospects of competing technologies in the local communications infrastructure of NGNs, the so-called next generation access networks (NGA), are not yet clear. In theory, Cable-TV networks, wireless technologies and in some regions even the copper cables can be upgraded to provide bandwidth that would substantially exceed current standards. It is still an open question what types of new applications will be developed for NGNs and what kinds of demand these applications will place on the bandwidth in the access network. Therefore, it is not yet decided which access technologies will be able to compete with FTTH or FTTC in quality and price. The problem for policy makers in regulating investments into FTTH and FTTC facilities in areas, in which they are considered to be monopolistic bottlenecks, is 122 Chapter 9 analyses how regulation of local access has responded to network convergence. 141 that, if these investments do not prove successful, the investing incumbent will be the one bearing the capital losses. However, if the investments are successful, then regulation allows competitors to share the facilities at regulated, cost-based access prices. Hausmann (1999) calls this a “free option” for competitors. The incumbent suffers the consequences of market uncertainties regardless of whether the price of capital goods is falling, whether demand is changing, or whether technological progress will make an investment obsolete, etc. Entrants do not bear any investment uncertainties or at least to a far lower degree (for the investments they make to interconnect with the incumbent’s investment). As a result, the level of investment by the incumbent is lower than would be economically efficient. Incumbents argue for “access holidays”, that is a temporary suspension of access regulation, such that their incentives for investment are increased (Heinacher and Preissl, 2006). They point out that the rents that can be earned during the regulation-free period will induce efficient investments. From the point of view of the disaggregated regulatory approach there can be no economic justification for access holidays (Knieps, 2006: 68f. and Knieps, 2005b: 88ff.). Access holidays can only be meaningful in cases where regulation would be justified due to ownership of monopolistic bottlenecks. And if such a bottleneck would be identified, so the disaggregated regulatory framework argues, regulation would be necessary from the very beginning. Otherwise consumers suffer a welfare loss from delayed competition in the market. To provide for investment incentives, the disaggregated approach implies that the regulatory contract needs to commit the regulator ex-ante to a proper compensation for the risks involved in making irreversible investments (Knieps, 2005b: 90).123 Blankart et al. (2007: 8) state that “the reference point for economically efficient investment signals is a market rate of return and not a monopolistic profit.” For practical regulatory purposes, Hausman (1999: 25f.) suggests calculating a mark-up on the TSLRIC (total service long-run incremental cost) price for assets affording irreversible investment using assumptions about the economic depreciation rate of capital, about the economic lifetime of an investment, a factor for changes in demand and changes in total factor productivity, and a variance for the stochastic process that determines the uncertainty. Since in the context of the disaggregated regulatory framework price-cap regulation is preferred over TSLRIC regulation, the risk compensation for new investments into irreversible assets should lead to a reduction of the factor x which reflects the expected productivity increase in the regulatory period. With the built-out of NGNs, regulators need to answer the question whether the extent of the monopolistic bottleneck identified in current PSTN network architecture is altered by the new network design. From the perspective of the disaggregated regulatory approach, the number of network elements, which competitors need access to, decreases once FTTH or FTTC are built out as a basis for offering VDSL 123 The Access Directive of the European Regulatory Framework (2002/19/EC, Art. 13(1)) explicitly allows taking account of investment risks in the price regulation of access. 142 connections to end-users. In the case of FTTC projects, for instance, the network upgrade consists of building fiber-optic cable closer to the customer premises. The incumbent draws a fiber-optic cable through the existing ductworks between the MDF site and the street cabinet. For this, the incumbent relies on ductworks and ducts that are part of the existing monopolistic bottleneck network infrastructure. It must be argued, that the extent of the monopolistic bottleneck network area is not increased by this investment of the incumbent, in the sense that entrants should be granted access to the new fiber cable. As long as both firms have the same opportunities for attracting customers in the retail market, the investment into additional fiber can be expected both of new entrants as well as of the incumbent. The prerequisite, that both have equal opportunities in the retail market presupposes that no entry barriers hinder competitors from entering the new market. Focusing again on long-run cost asymmetries between the incumbent and the competitors, it remains necessary that competitors are granted access to the ductworks and the ducts between the street cabinet and the main distribution frame as well as to the copper cable from the street cabinet to the customer premises for them to operate on the same cost function as the incumbent (Blankart et al., 2007: 14). Using these assets, the entrant can invest into fiber cable between the MDF site and the street cabinet at same costs as the incumbent. In the case of FTTH projects, competitors need equal access to the cable ducts which connect the end-user premises with the MDF site and to in-house cables. They can then draw their own fiber-optic cable to the enduser’s premises. To summarize, the extent of the monopolistic bottleneck network area will decrease in an NGN environment. When the metallic local loop is (partly) replaced with fiber-optic cables, then this is an investment, which can be undertaken by either the incumbent or its competitors. The ductworks and the cable ducts which connect end-user sites with higher-level network infrastructure remain monopolistic bottlenecks to which competitors must be granted equal access. Incumbents will face new regulation in that they must open their network infrastructure at points which were previously not regulated. 8.4 Conclusions This chapter clarified the importance of regulation in the Internet periphery for effective competition in Internet core markets by referring to the theory of marketpower leveraging. The principles of disaggregated regulation were presented. It was argued that minimally invasive regulation begins with a clear demarcation of the monopolistic bottleneck network area. The extent of the monopolistic bottleneck in local communications infrastructure was analyzed from the perspective of a regulator wanting to ensure non-discriminatory terms and conditions in broadband Internet access services. It was argued that in current PSTN networks competitors need to be able to co-locate their own transmission technology at the MDF site of the incumbent and need access to the bare copper cable leading to the end-user. In the future