In the Matter of ) ) Implementation of the Local ) CC Docket No. 96-98 Competition Provisions in the ) Telecommunications Act of 1996 ) REPLY COMMENTS OF THE NATIONAL TELECOMMUNICATIONS AND INFORMATION ADMINISTRATION Larry Irving Barbara S. Wellbery Assistant Secretary for Chief Counsel Communications & Information Shirl Kinney Phyllis E. Hartsock Deputy Assistant Secretary Deputy Chief Counsel Timothy Robinson Attorney Kathryn C. Brown Associate Administrator Mark Bykowsky Robert Cull Alfred Lee Tim Sloan Office of Policy Analysis and Development National Telecommunications and Information Administration U.S. Department of Commerce Room 4713 14th Street and Constitution Ave., N.W. Washington, D.C. 20230 (202) 482-1816 May 30, 1996
II. IMPLEMENTATION OF THE ACT'S INTERCONNECTION AND UNBUNDLING PROVISIONS
III. PRICING OF INTERCONNECTION AND UNBUNDLED NETWORK ELEMENTS
A. General Pricing Principles
B. LRIC: An Analysis
C. TSLRIC as Part of a Banded Approach
D. Allocating Costs Among Services and Interconnectors
In this rulemaking, the Commission will adopt rules to implement some of the most important parts of the Telecommunications Act of 1996 (Act) -- the provisions designed to foster competition in the market for local telecommunications services by requiring incumbent local exchange carriers (ILECs) (1) to interconnect with competing carriers and (2) to provide such carriers with unbundled network elements on just, reasonable, and nondiscriminatory terms. The Commission's decisions in this proceeding will determine, to a large degree, whether the nation realizes Congress' vision of a "pro-competitive, de-regulatory national policy framework designed to accelerate rapidly private sector deployment of advanced telecommunications and information technologies and services to all Americans."
NTIA recommends a dynamic national policy framework with respect to interconnection and unbundling that takes advantage of the combined experience of Federal and State regulators in matters of interconnection and unbundling and -- more fundamentally -- enlists the States in the implementation of the Act's provisions. Under this approach, the Commission would establish basic minimum interconnection and unbundling standards. The minimum set of unbundled network elements, for example, should include at least four network elements -- local loops, local switching, local transport and special access, and databases and signalling systems.
Furthermore, if (1) a State commission has ordered an ILEC to provide interconnection at a particular point or to unbundle a certain network element or (2) an ILEC has voluntarily offered to provide additional interconnection or unbundling, the Commission should create a rebuttable presumption that it is technically feasible for ILECs in any other part of the country to provide that same type of interconnection or that network element to any requesting carrier. An ILEC faced with such a request would have an opportunity to persuade the relevant State commission by clear and convincing evidence that offering the requested interconnection or network element is not technically feasible.
Every one or two years, moreover, the Commission should supplement its minimum interconnection/unbundling standards with additional requirements as a broader range of interconnection arrangements or network elements become more commonplace throughout the nation.
With respect to pricing, the Commission should establish a minimum set of principles governing the pricing of interconnection arrangements and unbundled network elements. Those principles define a "zone of reasonableness" within which negotiating parties may seek interconnection and unbundling rates. That zone should be defined at the bottom by TSLRIC. Its upper limit should be established through a "bottom-up" approach beginning at TSLRIC. Under this approach, an ILEC would have an opportunity to demonstrate by clear and convincing evidence that the TSLRIC estimate should be adjusted to include additional costs that: (1) contribute to a ILEC's long run average incremental cost of the services used by an interconnector and; (2) can be justified to the regulator as being clearly incremental to the provisioning of the service or functionality at issue, but is not part of the TSLRIC estimate for that service or functionality.
The Commission should also promulgate some basic rules for State commissions as they arbitrate contested negotiations. If one or more party requests State commission arbitration concerning the rate for a given interconnection arrangement or unbundled network element, the regulator should begin by defining, consistent with the Commission's pricing principles, the upper and lower bounds of the zone of reasonableness for that item. If both parties' proposed rates lie outside of the bound, the State commission should either send the parties back for further negotiation or choose a rate that falls within the zone. If one of the proposed rates is within the zone and the other is not, the regulator could either opt for the in-band price, or allow the party who lies outside the band to make a final offer within the zone. If both offered prices are within the zone, the Commission should require that the State regulator choose either of the two prices.
The National Telecommunications and Information Administration (NTIA), an Executive Branch agency within the Department of Commerce, is the President's principal advisor on domestic and international telecommunications and information policy. NTIA respectfully replies to comments submitted in response to the Commission's Notice of Proposed Rulemaking (Notice) in the above-captioned proceeding.y
In this rulemaking, the Commission will adopt rules to implement some of the most important parts of the Telecommunications Act of 1996 (Act)y -- the provisions designed to foster competition in the market for local telecommunications services by requiring incumbent local exchange carriers (ILECs) (1) to interconnect with competing carriers and (2) to provide such carriers with unbundled network elements on just, reasonable, and nondiscriminatory terms. Other sections of the Act contemplate additional entry into markets where competition is already considerable (for example, Bell Operating Company (BOC) entry into long distance, manufacturing, and information services). In contrast, the local competition provisions -- principally Sections 251 and 252y -- are designed to stimulate entry in areas where competition is, at best, nascent and, thereby, create a process that over time will give consumers service choices where they now have few or none. The Commission's decisions in this proceeding will determine, to a large degree, whether the nation realizes Congress' vision of a "pro-competitive, de-regulatory national policy framework designed to accelerate rapidly private sector deployment of advanced telecommunications and information technologies and services to all Americans."y
There are many potential benefits of a national framework governing interconnection and network unbundling (including the prices at which those arrangements are offered).y National rules will give negotiating parties greater certainty as to their rights and obligations under the Act.y This could, in turn, speed the bargaining process by eliminating possible areas of dispute,y and by reducing the possibility that prospective entrants may have to renegotiate or relitigate the same issues in multiple States.y National requirements also could help new entrants to develop rational business plans and network deployment schedules, thereby facilitating competitive entry and limiting its potential costs.y Finally, such rules could expedite arbitration and review of interconnection agreements by the Commission, State regulators, and the courts.y
To be fully successful, however, any national framework for interconnection and unbundling must give private parties sufficient latitude to conduct and conclude the negotiations that Congress intended would drive the development of local telephone service competition. There is, in all likelihood, a wide range of interconnection agreements that can effectively satisfy the needs of the parties to those agreements and the interests of their current and prospective customers. The national framework adopted in this proceeding must give private firms a full opportunity to pursue such agreements, while creating incentives for them to negotiate in good faith and to reach mutually acceptable outcomes.
The national framework should also foster a dynamic environment in which the Commission and State commissions act collaboratively to achieve the Act's goals. The Act entrusts the Commission with important new responsibilities to ensure that all U.S. telecommunications markets are opened to competition. But the Act gives State commissions the lead in supervising (and, in some cases, concluding) the private interconnection agreements that will give rise to local competition. State regulators have gained valuable experience and insights as they have worked to create entry opportunities for new providers of local telecommunications services, particularly concerning the types of interconnection arrangements that are both feasible and necessary to make local entry viable.
The Commission should therefore develop a policy framework that takes advantage of the combined experience of Federal and State regulators in matters of interconnection and unbundling and -- more fundamentally -- enlists the States in the implementation of the Act's provisions. By so doing, the Commission will ensure that its new regulations not only satisfy the Act's procedural timetables, but also create a dynamic process that can help achieve the Act's central objective -- meaningful and lasting competition in the local telecommunications service market.
II. IMPLEMENTATION OF THE ACT'S INTERCONNECTION AND UNBUNDLING PROVISIONS
Section 251 imposes three fundamental obligations on ILECs:y (1) to interconnect with other telecommunications carriers at any "technically feasible point" on just, reasonable, and nondiscriminatory terms;y (2) to afford requesting carriers access to network elements on an unbundled basis at any technically feasible point on just, reasonable, and nondiscriminatory terms;y and (3) to offer for resale at wholesale rates any telecommunications service made available at retail.y The Act charges the Commission with establishing regulations to implement those requirements.y
In the Notice, the Commission solicits views on the nature of those implementing regulations. In particular, it seeks comment on the wisdom of adopting "explicit," "uniform," "specific" national rules concerning an ILEC's interconnection and unbundling responsibilities.y In response, a number of parties have urged the Commission to designate a detailed and fixed minimum set of technically feasible interconnection points and unbundled network elements. AT&T and MCI, for example, have advocated that the Commission require the unbundling of 11 and 13 specific network elements, respectively.y
Notwithstanding the potential benefits of such an approach, NTIA recommends a national framework that is more dynamic and expansive in approach. As the Commission recognizes, detailed Federal standards could actually limit States' ability to experiment with alternative, yet nonetheless pro-competitive, approaches to interconnection and unbundling or to broker interconnection agreements that better match local technological, geographic, or demographic conditions.y
As importantly, Commission-established static, minimum requirements, even if periodically reviewed and revised,y will almost by their very nature be underinclusive, because at any point in time the Commission cannot anticipate all possible forms of interconnection and unbundling. The potential adverse effects of a static approach will become particularly acute in future years as the range of feasible interconnection points and unbundled network elements expands and the benefits of local competition become more apparent and more widely enjoyed.
There is the further risk that any fixed "floor" for interconnection and unbundling would become a ceiling as well. Given that some ILECs may not be eager to complete interconnection agreements that will begin or speed the erosion of their local service monopolies,y they may use the Federal minimums as a shield to ward off requests for additional interconnection/unbundling. In such cases, further progress could be stymied until the Commission completed proceedings to adopt additional requirements.
To avoid these potential difficulties, NTIA recommends that the Commission adopt a more dynamic approach to interconnection and unbundling, similar to that adumbrated in certain parts of the Notice.y This approach would operate as follows: First, the Commission would establish basic minimum interconnection and unbundling standards. In this regard, NTIA agrees with the Commission that, with respect to unbundling, the Act requires ILECs to provide reasonable and nondiscriminatory access to at least four network elements -- local loops, local switching, local transport and special access, and databases and signalling systems.y
Further, the Commission effectively has mandated the unbundling of special access and local transport in its Expanded Interconnection proceeding.y There is no reason to reverse that decision. The Commission also has concluded tentatively, based on decisions by several State commissions, that it is technically feasible and desirable to unbundle further local loops.y Given the Act's strong bent towards competition, NTIA believes that the impetus should be in favor of more unbundling, not less. There is, however, enough variability among ILECs' networks and different perceptions among the States about the feasibility of opening those networks that it will be difficult for the Commission to make definitive judgments about the appropriateness of imposing numerous additional requirements at this time. NTIA believes that these tensions can be reconciled by an approach centered on State-by-State determinations, guided by national pro-competitive principles.
This brings us to the second component of NTIA's proposal. If (1) a State commission has ordered an ILEC to provide interconnection at a particular point or to unbundle a certain network element or (2) an ILEC has voluntarily offered to provide additional interconnection or unbundling, the Commission should create a rebuttable presumption that it is technically feasible for ILECs in any other part of the country to provide that same type of interconnection or that network element to any requesting carrier.y An ILEC faced with such a request would have an opportunity (in the event that the parties did not reach a mutually acceptable resolution of the matter) to persuade the relevant State commissiony by clear and convincing evidence that offering the requested interconnection or network element is not technically feasible.y By allowing for the possibility that actions in one State will create interconnection and unbundling opportunities in other States that exceed the national minimums, without Commission intervention, this approach will substantially mitigate concerns that those minimums could also become maximums.
NTIA expects proceedings at the State level to make available a steadily increasing range of interconnection points and unbundled network elements available to competitive entrants
in a growing number of jurisdictions.y The Commission would further this process by periodically supplementing its minimum interconnection/unbundling standards with additional requirements as a broader range of interconnection arrangements or network elements become commonplace throughout the nation. States have accomplished much already with respect to interconnection and unbundling and there will probably be many private negotiations concluded in the coming months. NTIA therefore believes that the Commission's first review of its minimum interconnection and unbundling requirements should occur no later than one year after it promulgates rules in this proceeding. Subsequent reviews should then be conducted every one or two years thereafter.
Under the foregoing approach, the Commission could ensure that new entrants benefit from the development of a continually-increasing, common set of interconnection and unbundling arrangements throughout the country, without having to engage in lengthy and potentially costly negotiations.y At the same time, States would be permitted, indeed encouraged, to mandate additional unbundling and interconnection requirements based on particular local market conditions and the agreements concluded by private parties. Thus, NTIA's framework would create a partnership between Federal and State regulators that would allow them to bring their combined expertise and experience to bear on the question of how interconnection and unbundling standards or requirements can best be used to foster increased competition in the provision of local telephone services.
NTIA's proposed approach is consistent with the letter and spirit of the Act because it involves the Commission, State commissions, and private negotiators in defining an ILEC's interconnection and unbundling obligations. Moreover, in expanding the number of actors involved in determining such obligations, NTIA's proposal ensures that the process will be dynamic and not unduly affected by the decisions of a single entity. Although NTIA's approach will not eliminate local or, perhaps, regional variation in the interconnection/unbundling requirements prescribed, it will create a trajectory of increasing interconnection and unbundling obligations, with all of the competitive benefits that such a path is likely to generate.
III. PRICING OF INTERCONNECTION AND UNBUNDLED NETWORK ELEMENTS
If Section 251's interconnection and unbundling provisions are designed to promote competition, the pricing of network interconnection and unbundled network elements will determine to a large extent how effective Section 251 will be in achieving that objective. Generally speaking, the Act requires that the rates established for interconnection or for network elements should be: (1) nondiscriminatory and (2) based on the costs of providing such interconnection or element, including a reasonable profit.y The Notice seeks comment on how best to ensure compliance with that standard.
NTIA agrees with the Commission and the Department of Justice that the Commission should establish a minimum set of principles governing the pricing of interconnection and unbundled network elements.y As the Commission points out, Section 251(d) of the Act requires the Commission to adopt regulations to implement the statutory requirements for ILECs to provide interconnection and unbundling on "just, reasonable, and nondiscriminatory" terms.y Nationwide pricing principles would furnish a certain and consistent benchmark for assessing ILECs' compliance with that fundamental obligation.
Commission-prescribed pricing principles would also serve the Act's overriding pro-competitive goals in the same way as a national framework with respect to interconnection and unbundling. They would, for example, increase the predictability of interconnection and unbundling pricing, making it easier for new entrants to develop their business plans and facilitating negotiation, arbitration, and review of interconnection agreements between entrants and ILECs.y At the same time, national pricing principles would not encroach upon the rights of State commissions because they would retain responsibility for applying those principles (in contested negotiations) to establish or to approve specific prices in particular circumstances.y
We should emphasize here that any pricing principles adopted by the Commission should give negotiating parties every opportunity to reach mutually acceptable agreements with respect to price as well as other network interconnection and unbundling details. In the event that negotiating parties fail to reach agreement, the State regulator should establish the appropriate price for the interconnection to be provided, based on specific Commission-prescribed guidelines that embody the pricing objectives described below.
A. General Pricing Principles
In establishing national principles for pricing interconnection arrangements and unbundled network elements, the Commission's central objective should be that identified by Congress -- to ensure that the rates for those offerings reflect underlying costs. Regulations that govern telecommunications service pricing should be designed to produce service prices equal to the opportunity costs society incurs from having its scarce resources devoted to the production of these services.y Policies that deviate from this basic principle will send incorrect price signals and thereby distort ILEC and new entrants' investment decisions, as well as consumer purchases, in a manner that harms society's economic welfare.y
On the one hand, requiring ILECs to establish prices that are below the true costs of providing their services will likely encourage excessive entry by resellers in the short run, thereby expanding service consumption beyond the economically efficient point. Below-cost pricing also may not allow ILECs to make economically efficient investments in telecommunications infrastructure. On the other hand, allowing ILECs to price services above costs will retard entry by resellers, depress service consumption relative to the economic optimum, and invite economically inefficient facilities-based competitive entry.
Establishing a set of economically efficient prices is complicated by at least two factors. First, provision of telecommunications services appears to be characterized by significant joint and common costs.y There is no easy way to apportion such "shared" costs among the services involved so as to establish the "true" costs of providing each service. Second, setting efficient prices requires considerable information on underlying costs that are typically within control of the serving ILEC, rather than the relevant regulator. The ILEC obviously has no strong incentive to reveal private cost information that may do it harm.
B. TSLRIC: An Analysis
Given the asymmetry in cost information between regulators and regulated firms, the Commission should not sanction any costing methodology that relies heavily on ILEC-provided information. A number of commenters in this proceeding have proposed an alternative approach, referred to as total service long run incremental cost (TSLRIC). Because TSLRIC cost estimates are not based on ILEC-provided information, NTIA believes that they can be very useful to the Commission and State regulators in securing efficient and pro-competitive prices for interconnection and unbundled network elements. As discussed below, TSLRIC can be particularly valuable if employed as a lower bound with which to structure negotiations.
In their most basic form, however, TSLRIC estimates alone may not be sufficient to accomplish the regulator's pricing objective -- fostering competition through interconnection rates that accurately reflect input costs.y For instance, as NARUC correctly notes, "because some shared costs are not included in the calculations of TSLRIC, a firm that collects only TSLRIC from all of its services will not be able to cover its total costs, including overheads, and could not remain viable over the long term."y,y
Further, estimation of TSLRIC requires the identification of a cost-minimizing service output level. This, of course, requires a prediction regarding the service's peak load (or busy-hour) demand. TSLRIC assumes no change in the size of the service's peak load demand. That assumption can be criticized in several respects. Today's demand figures result from a pricing structure that should change with the arrival of competition. More importantly, current demand is based on service prices that may be substantially higher than service prices under TSLRIC. Unless demand for network elements is perfectly inelastic, any lower prices will stimulate, perhaps substantially, service demand.y Using TSLRIC estimates alone may therefore underestimate the ILEC's true cost of creating and maintaining a network of sufficient size.y
Assuming one could overcome these difficulties, relying solely on TSLRIC-based prices may be unwise for several reasons. Although interexchange carriers (IXCs) have argued that a forward-looking, least cost methodology reflects the "true" economic cost of providing service, TSLRIC estimates differ from standard economic long run costs. Typically, a firm's costs reflect not the plant it could conceivably create, but rather the plant it already has. The theory of long run adjustment in a competitive market holds that, if the provider's plant does not permit the most efficient means of production, potential entrants have an incentive to construct the appropriate plant, implement the least cost production process, and undersell the incumbent. In the long run, only those firms that use the least cost technology will remain in business.y
Further, to the extent that these rates reflect a low cost network that is yet to be created, incumbents may have to accept interconnection payments below their actual costs of supplying service. Also, if interconnectors receive rates associated with the least cost method of production, none will have the incentive to create its own efficient network. Finally, no service provider would have an incentive to operate the existing network if constrained to provide interconnection at TSLRIC-based rates that are below average cost.
One important potential advantage of the TSLRIC approach, however, is its relative ease of calculation. Rather than estimate costs reflecting the present ILEC network -- a difficult task even if ILECs provided reliable data -- it is possible to generate TSLRIC estimates based on a "green field" approach, which assumes construction of a network from scratch. A recent TSLRIC model created by Hatfield Associates (Hatfield),y for example, estimates the least cost method of providing a service to a given number of customers with the best available technology. To its credit, Hatfield's methodology employs a forward-looking cost approach. But, a TSLRIC model based solely on "green field-derived" costs may not measure forward-looking costs sufficiently from the perspective of either the ILEC or the regulator. The ILEC's forward-looking costs are reflected in its short-run marginal cost curves -- not the cost point estimate generated by Hatfield's green field methodology.y
Despite the limitations of a TSLRIC approach, it nevertheless offers an objective, tractable means of estimating the costs of an ILEC's services and functionalities. For that reason, NTIA believes that TSLRIC should be part of the pricing principles that the Commission adopts to ensure just and reasonable rates for ILEC interconnection arrangements and unbundled network elements.
C. TSLRIC as Part of a Banded Approach
The optimum costing methodology would allow the Commission and State commissions to set interconnection rates at the long run average incremental cost (LRAIC) of service for the present network.y In this way, incumbents would recover their economic costs of production for the service in question,y including capital costs,y without also recouping costs incurred solely in the provision of other services. NTIA readily acknowledges, however, that given the informational asymmetry between ILECs and regulators, it may be difficult to calculate LRAIC accurately.
For that reason, we believe that the Commission should establish pricing principles that define a "zone of reasonableness" within which negotiating parties may seek interconnection and unbundling rates. That zone should be defined at the bottom by TSLRIC.y Its upper limit should be established through a "bottom-up" approach beginning at TSLRIC.y An ILEC would have an opportunity to demonstrate by clear and convincing evidence that the TSLRIC estimate should be adjusted to include additional costs -- joint, common, or othery -- that: (1) contribute to a ILEC's long run average incremental cost of the services used by an interconnector and; (2) can be justified to the regulator as being clearly incremental to the provisioning of the service or functionality at issue, but is not part of the TSLRIC estimate for that service or functionality.y
The above procedure leaves open the question of whether, and if so how, to recover any differences between historical costs and TSLRIC's forward looking costs. Clearly one should consider for recovery only those costs that the ILEC can demonstrate convincingly to the regulator are incurred in service provisioning. Moreover, clearly the ILEC may not be entitled to recover any or all of its embedded costs, no matter how prudently incurred. First, the regulator should require the ILEC to mitigate these costs in instances where greater efficiency can be gained through the existing network. Secondly, any additional benefits that the ILECs gain as a result of changes in the regulatory environment should be taken into account (e.g., benefits from interLATA entry). There may be some remaining shortfall, but its size, of course, will not be determined for a number of years, certainly until after interLATA entry. As a result, if the shortfall is to be recovered at all it would be unwise to include it in input prices before the size of the shortfall is better known. Finally, we note that if there are any historical costs that should be recovered, it is preferable to recover them through input prices as a last resort only. Recovery through input prices should be done only through a competitively neutral manner.
The price negotiated by the parties could fall anywhere within that zone of reasonableness.y NTIA expects that the parties will bring to the negotiation their assessments of how the relevant State commission will apply the Commission's pricing principles to define the zone in particular circumstances.y Each party's assessment of the State regulator's likely decision will be an important component of its bargaining strategy, because that assessment will establish the odds of having unfavorable pricing terms imposed upon them. In this way, Commission-established principles for establishing the zone of reasonable prices would increase the parties' incentives to negotiate in good faith.y
The Commission can strengthen those incentives by promulgating some basic rules for State commissions as they arbitrate contested negotiations. If one or more party requests State commission arbitration concerning the rate for a given interconnection arrangement or unbundled network element, the regulator should begin by defining, consistent with the Commission's pricing principles, the upper and lower bounds of the zone of reasonableness for that item. If both parties' proposed rates lie outside of the bound, the State commission should either send the parties back for further negotiation or choose a rate that falls within the zone.y If one of the proposed rates is within the zone and the other is not, the regulator could either opt for the in-band price, or allow the party who lies outside the band to make a final offer within the zone.y If both offered prices are within the zone, the Commission should require that the State regulator choose either of the two prices.y
D. Allocating Costs Among Services and Interconnectors
As alluded to above, assigning costs to individual services or functionalities is difficult in the presence of substantial shared costs. One way to handle this problem is to compute a percentage mark-up on the LRIC of each service, as Hatfield does. Although Hatfield employs a six percent mark-up, that estimate may not be reasonable for all ILECs. If the size of total shared costs is known, one could conceivably calculate the unique percentage mark-up applied to all services such that shared costs are just covered. In practice, it would be wise for the State regulator to employ a relatively low estimate of shared costs in the lower bound of its price band for a service (e.g., Hatfield's), a higher estimate in the upper bound, and then permit the parties to arrive at a mutually beneficial arrangement through negotiation.y
Another potential problem arises in apportioning LRIC among users (both interconnectors and the incumbent ILEC) who share a facility. The TSLRIC methodology, for example, derives cost estimates for an entire service or functionality; it does not specify how that total cost should be apportioned among users.y The costs of a shared facility should be apportioned based on the capacity required by the individual interconnectors:
Pricing would be capacity based if a user paid at each point of time in relation to the depreciation charges for that part of the capacity that, either at the time of investment, or at the latest revaluation of the asset because of changing market conditions, was "reserved" for him. . . . It would. . . not be unrealistic to apply capacity-based pricing to large users, in particular interconnecting network operators. Such users have predictable consumption patterns. . . . Capacity-based costing would not preclude charging for operating costs on the basis of actual usage with which this kind of cost varies. Nor would it preclude, of course, charges based on actual usage if the latter exceeds the capacity that was reserved for the demander.y
In New York, parties have negotiated capacity-based pricing arrangements that employ measures of actual usage. Under the NYNEX-Teleport agreement:
The general form of the agreement is to establish a particular charge for a two-way channel of given capacity between the two companies. Traffic is measured at the busy hour each month and the relative measurements are used as an allocation factor for the established channel rate. If traffic is exactly balanced, the payments to each company cancel out and the level of the established rate is irrelevant. If traffic is not balanced, and if Teleport, for example, sends more traffic to NYNEX than it receives from NYNEX at the busy hour, that imbalance is used to compute a net payment from Teleport to NYNEX.y
As Brock and others have noted, the distinction between peak and off-peak traffic (1) provides an administratively simple way to calculate contributions from firms and (2) promotes economic efficiency because costs are "generally associated with peak traffic and therefore the effectively zero charge for terminating off-peak traffic is cost based."y Again, the exact method for sharing costs among users should be agreed upon through negotiation and may vary across markets. The NYNEX-Teleport agreement is but an example of a usage-based agreement that promotes economic efficiency.
For the foregoing reasons, NTIA respectfully requests that the Commission adopt the recommendations contained herein.
Larry Irving Assistant Secretary for Barbara S. Wellbery Communications & Information Chief Counsel Shirl Kinney Deputy Assistant Secretary Phyllis E. Hartsock Deputy Chief Counsel Timothy Robinson Attorney Kathryn C. Brown Associate Administrator Mark Bykowsky Robert Cull Alfred Lee Tim Sloan Office of Policy Analysis and Development National Telecommunications and Information Administration U.S. Department of Commerce Room 4713 14th Street and Constitution Ave., N.W. Washington, D.C. 20230 (202) 482-1816 May 30, 1996
 Implementation of the Local Competition Provisions of the Telecommunications Act of 1996, CC Docket No. 96-98, FCC 96-182 (released Apr. 19, 1996) (Notice). Unless otherwise indicated, all subsequent citations to "Comments" shall refer to pleadings filed on May 16, 1996 in CC Docket No. 96-98.
 Telecommunications Act of 1996, Pub. L. No. 104-104, 110 Stat. 56 (Act).
 For convenience, all references to the Act in this pleading will cite to the section numbers that will apply after the Act's provisions have been codified in the United States Code.
 H.R. Conf. Rep. No. 104-458, 104th Cong., 2d Sess. 1 (1996), reprinted in 1996 U.S.C.C.A.N. 124 (Joint Explanatory Statement).
 Collocation is clearly part and parcel of any discussion of the term "interconnection." Accordingly, NTIA's proposed national framework for interconnection applies to collocation as well.
 NTIA believes that these negotiations ought to establish not only network interconnection and unbundling arrangements and the price terms for those arrangements, but also self-executing penalties for failure to comply with a party's provisioning and performance obligations under an agreement. See Comments of the Association of Local Telecommunications Services.
 Notice 50. See also Comments of the United States Department of Justice at 9-12 (DOJ Comments).
 DOJ Comments at 12.
 See, e.g., Notice 30, 50, 79. See also DOJ Comments at 13.
 Notice 31, 32; DOJ Comments at 12, 14.
 The term "ILEC" includes the BOCs and GTE, which together serve about 86 percent of the nation's 157.9 million access lines. See United States Telephone Ass'n, Phone Facts 1995, at 3, 10. It also encompasses the more than 1300 other companies that have historically provided local telephone service to the remainder of the country. Act 251(h)(1). The Act empowers the Commission to classify other telecommunications service providers with certain characteristics as ILECs. Id. 251(h)(2). At this time, however, no alternative provider appears to meet the statutory criteria for being categorized as an ILEC. Until a competing carrier has been classified as an ILEC, it should not be subject to the additional obligations that the Act imposes only on ILECs. See Notice 45; DOJ Comments at 22-23.
 Act 251(c)(2).
 Act 251(c)(3). As the Commission points out, Congress clearly intended that the term "network element" includes both "'a facility or equipment used in the provision of a telecommunications services' as well as 'features, functions, and capabilities that are provided by means of such facility or equipment.'" Notice 83 (quoting Act 153(29) and Joint Explanatory Statement, supra note 4, at 116).
 Id. 251(c)(4). An ILEC has a general duty to negotiate in good faith to reach agreements that particularize those basic three obligations. Id. 251(c)(1). NTIA believes that the responsibility to bargain in good faith means, at a minimum, that an ILEC must enter into negotiations without preconditions. As the Notice indicates, some ILECs have allegedly refused to commence talks "until the requesting carrier satisfies certain conditions, such as signing a nondisclosure agreement, or agreeing to limit its legal remedies in the event that negotiations fail." Notice 47. The Commission should state unequivocally that any attempt to impose such preconditions will be considered a breach of an ILEC's duty to negotiate in good faith.
 Act 251(d)(1). The Commission asks whether its rules should apply to both interstate and intrastate aspects of interconnection, unbundling, and resale. Notice 37-39. Section 251(c)(2) obligates ILECs to interconnect with other carriers so that the latter can offer, among other things, telephone exchange service -- the quintessential intrastate offering. Similarly, Section 251(c)(4) requires ILECs to permit resale of any of their telecommunications services, not just their interstate ones. In short, Congress plainly intended that the Federal implementing regulations commanded by Section 251(d) would affect intrastate as well as interstate services.
Furthermore, the structure of Section 252, which mandates carrier-to-carrier negotiations and government oversight of those negotiations, makes no sense if the obligations of Section 251 are limited to interstate services. Section 252 gives State commissions the primary role in supervising the bargaining process, subject to Federal guidelines. If the negotiations concerned interconnection, unbundling, and resale solely with respect to interstate services, a State role would be unnecessary and inappropriate because regulation of those services have always been the exclusive province of the Commission. Congress' blueprint of joint Federal and State action in this area evinces its expectation that an ILEC's interconnection, unbundling, and resale responsibilities under Section 251 would extend to both their interstate and intrastate offerings.
 See Notice 27, 50-51.
 See id. 32.
 Id. 33. Although Congress has expressed its commitment to a national policy framework, the procedures that Congress established for concluding interconnection agreements will tend toward variation in outcome, rather than uniformity. As noted above, the quest for such agreements will begin with voluntary negotiations between myriad private firms. If the parties cannot agree, the scene shifts to State regulatory commissions for arbitration with, perhaps, subsequent review by one of many Federal courts. See Act 252(a), (b), (e)(6). In instances where a State commission refuses to arbitrate a disagreement between the parties to a negotiation, the Commission may assume jurisdiction in the State's place. Id. 252(e)(5). That process clearly does not contemplate establishment of a single model interconnection agreement.
Court review of disputed State-approved interconnection agreements might provide an opportunity for the Commission to impress some consistency on the results of that process. The Act provides that, upon review of State-approved agreements, a Federal court must "determine whether the agreement . . . meets the requirements of section 251 and [section 252]." Id. 252(e)(6). As the expert agency charged with administering the Act, the Commission would seem best equipped to assist the court in making that determination. Thus, the Commission should seek to intervene in any such appeal.
 See Notice 57.
 See, e.g., DOJ Comments at 9-10. The Act assumes that the prospect of interLATA entry will attract the BOCs to the bargaining table. That carrot will obviously be ineffective against the hundreds of ILECs that are presently free to provide long distance services. Furthermore, it is far from clear that all of the BOCs value the provision of interLATA services more highly than the protection of their lucrative local service markets. See, e.g., Telecommunications Reports, May 20, 1996, at 12 (detailing Southwestern Bell's attempts to overturn a Texas Public Utilities Commission order facilitating competitive local entry by Teleport Communications). Finally, interLATA entry remains an inducement, if at all, only so long as entry is denied. Once a BOC gains authority to provide long distance services, it may reassess its willingness to negotiate additional interconnection agreements.
 E.g., Notice 29, 58-59, 80.
 See id. 94, 98, 104-105, 107.
 Id. 104.
 Id. 97.
 The presumption should apply not only to the availability of facilities, but also to provisioning arrangements, such as installation, maintenance, repair, and the imposition of non-recurring charges. See id. 61.
As indicated in the Notice, a number of States have already ordered or are considering whether to order further unbundling of switching, signalling, and database functions beyond the level required by the Act or proposed by the Commission. Id. 100, 109. Under NTIA's proposal, a prospective entrant in another State could rely on those decisions to request similar subelements from the serving ILEC. This provides a clear example of how a policy framework that enlists State commissions in the determination of appropriate interconnection and unbundling requirements could produce more interconnection and unbundling than one that relies solely or primarily on Commission action.
 The Commission would, of course, arbitrate the disagreement in the event that the State commission declined to do so.
 Cf. Notice 58 (suggesting that in the event of a dispute over interconnection, "the incumbent LEC has the burden of demonstrating that interconnection at a particular point is technically infeasible").
In addition, the Commission correctly concludes that any party alleging harm to the network from the provision of requested interconnection points should be required to present detailed information to support claims of network harm. Id. 56.
 As States and negotiating carriers move beyond the minimum requirements, NTIA expects that the industry will and should step forward to test and to standardize additional interconnection points and unbundling arrangements as they become more commonly available in a growing number of markets.
 As networks evolve due to interconnection/unbundling agreements and other improvements, strong and effective network and information disclosure rules will play an even more important role in promoting the development of competition. See Notice 189-193.
 Act 252(d)(1). The Act's specification that rates may provide a reasonable profit is somewhat redundant because conventional economic measures of "cost" assume a reasonable or "normal" profit.
 Notice 39-40; DOJ Comments at 24-26.
 Notice 117 (citing Act 251(c)(2), (c)(3)).
 DOJ Comments at 25; Notice 40.
 DOJ Comments at 25.
 Prices should be based upon "forward looking" costs -- the expenditures that a business will incur or avoid as a result of future decisions. "Economically efficient pricing looks not to the past -- not to how we got where we are -- but to the future; efficiency requires that prices tell customers what incremental resources society will use if they take more of the good or service in question, what resources society will save if they consume less of it." Alfred Kahn and William Shew, Current Issues in Telecommunications Regulation: Pricing, 4 Yale J. on Reg. 191, 224 (1987) (Kahn and Shew).
 The extent of this "allocative" distortion depends upon the level of usage and investment either stimulated or stifled by inefficient pricing.
 See, e.g., Kahn and Shew, supra note 35, at 194. Joint costs are expenditures that are incremental to a group of services, but not to any particular service within that group. Common costs are expenditures that, at least in theory, can be attributable to a particular service.
 Cf. DOJ Comments at 30 ("If the prices that competing carriers pay for their inputs are distorted from the true economic costs of those inputs, those prices will lead carriers to choose technologies that will minimize their use of overpriced but lower cost inputs, thus impairing the quality or increasing the cost of their service offering"). This implies that if prices are below the "true economic costs" of inputs, carriers will choose technologies that maximize their use of underpriced but higher cost inputs.
 See id. at 35. Hatfield's TSLRIC model does attempt to account for some administrative overhead:
Certain costs that vary with the size of the firm, and therefore do not meet the economist's definition of overhead, are often included under the classification of General and Administrative expenses. For example, if a LEC did not provide loops, it would be a much smaller company, and would therefore have lower costs. Some of those costs are attributed to overhead under current LEC accounting procedures. We therefore include a portion of these "overhead" costs in our TSLRIC estimates.
Historical overhead expenses for the LECs, such as administration, planning, legal, and human resources seem excessive when compared to firms that operate in a competitive environment. The relationship between revenues and overhead for selected firms in the auto manufacturing and airline industries was examined. A six percent overhead loading factor was found for these industries. The cost of the functions that this factor is used to estimate should not vary widely across industries. In other words, the relationship between revenues and administration, planning, legal, and human resources are likely to be similar in the telecommunications industry.
Hatfield Associates, Inc., The Cost of Basic Network Elements: Theory, Modeling and Policy Implications 30 (March 1996) (prepared for MCI Telecommunications Corp.)
We note that by defining shared costs as not being incremental to any one service, we do not imply that these costs are invariant to the size of the firm, as Hatfield contends. It is also unclear how sensitive Hatfield's TSLRIC estimates are to changes in the six percent overhead factor. To the extent that Hatfield excludes some shared costs or its six percent factor underestimates true overhead costs in telecommunications, its TSLRIC model may underestimate the ILECs' costs.
 Economists generally agree with NARUC's assessment that TSLRIC prices do not cover all of the ILECs' joint and common costs. For example, Harris and Yao note that
In addition to TSLRIC, LECs also have shared costs which are incurred for facilities and resources used in the production of two or more services, and can therefore not be eliminated by the discontinuation of a single service. Examples of shared costs include fiber strands used for transport services, and stand-by modular switching capacity. These shared costs are incurred whenever LECs provide services to end-users and should therefore be reflected in retail, wholesale, unbundled network element and call termination prices. Some portion of common costs also need to be recovered. Common costs are incurred through facilities and resources used in the production of all the LECs services.
Robert Harris and Dennis Yao, Federal Implementation of the Telecommunications Act of 1996: Competition in the Local Exchange 18 (May 1996) (Attached to Comments of US West, Inc.) (Harris and Yao). Harris and Yao used the term shared costs in the way that we (and NARUC) use the term joint costs. We use the term shared cost to refer to both joint and common costs.
Potential interconnectors acknowledge that joint and common costs are part of the long-run incremental costs of a service that should be recovered. They are legitimately concerned nevertheless, about an IEC's' incentives to overstate those costs. As AT&T puts it:
the TSLRIC of a network element includes all of the element-specific investment needed to construct and operate the facilities used to produce that element, including costs that are fixed in the short run. There may well be some cases of non-trivial "common" or "shared" costs, however, and, particularly in light of the potential for confusion and abuse in this area, it is critical that the Commission establish rules to constrain the ILECs' incentives and abilities to manipulate the quantification and allocation of "common" or "shared" costs in ways that thwart competition.
Comments of AT&T Corp. at 62-63 (emphasis in original) (footnotes and citations omitted). NTIA strongly agrees that only those costs that are truly incremental to the provision of a service or functionality should be considered in reviewing interconnection rates.
 The existence of economies of scale and peak-load demand problems in the telecommunications sector suggests that identifying economically efficient prices requires not only cost studies, but analyses of service demand as well. Because TSLRIC largely ignores the service demand dimension, the use of TSLRIC alone will be limited in its ability to yield an economically efficient set of prices.
 A second drawback of TSLRIC is that it is based on the cost of providing an entire service, as opposed to traditional economic cost measures that identify the cost of providing additional units of a service (e.g., LRIC). In practice, however, it may not be possible to measure accurately changes in average costs associated with changes in output. As a result, the best available "increment" for deriving incremental costs will likely be the entire service or functionality. Indeed, the cost principles we advocate below employ a service or functionality as the increment.
 In this respect, TSLRIC runs up against the theory of long-run adjustment in a competitive market:
Even when they do not have any shared and common costs, firms in competitive industries experiencing rapid technological change do not price their goods and services at TSLRIC. Under standard economic theory, the least efficient, viable producer in an industry would earn zero economic profits. All producers using older technology are, at least in the short-run, forced to either upgrade their plant or exit the market. Conversely, positive economic profits are earned by the most efficient and innovative firms in a competitive market. Therefore, what often occurs in competitive industries is that a production facility makes above average profits during its early years of operation which decline over time until the firm is forced to upgrade or close down the production facility. However, if industry-wide prices were set at TSLRIC, only the most efficient producer using the latest technology would be able to cover its costs and make a profit.
Harris and Yao, supra note 40, at 19.
 Hatfield Associates, Inc., The Cost of Basic Network Elements: Theory, Modeling and Practice (Mar. 1996) (prepared on behalf of MCI Telecommunications Corp.).
 We understand that the Hatfield model has been revised to address concerns about its green field nature. Specifically, it appears that the approach has evolved from a "scorched earth approach [that assumes] no direct interoffice trunking and homogeneous population density zones" to a "scorched node approach using outputs from the Benchmark Cost Model [developed cooperatively by IXCs and ILECs] to size loop plant requirements." Comments of AT&T Corp., App. E at 1 n.1. Because we believe that TSLRIC estimates should be based on actual ILEC costs, we support, as a general matter, any attempt to modify the Hatfield model or any other TSLRIC model in that direction.
 As the Department of Justice notes:
The TSLRIC methodology would price network elements at the long-run, forward-looking economic costs of the particular network element, given the efficient provision of all other network components by the ILEC. This standard would be "forward looking" in that it would be based on the best generally available technology, current input prices, and economic cost-minimization. It would be "long run" in that it would include the forward looking capital costs necessary to provide the element. It would define and utilize the network element as the appropriate "increment," and its added cost would be the added economic cost of the element conditioned on the provision of other network components.
DOJ Comments at 27. NTIA's LRAIC also uses the network element as the appropriate increment, although NTIA recognizes that the total costs of the network must be apportioned among all users. The average cost measure makes it easier to accomplish this; by multiplying average cost by some measure of an interconnector's use of the element (e.g., number of subscribers or calls), the interconnection cost for each user can be calculated. We discuss this in more detail below. NTIA's LRAIC is long run in the sense that it includes the capital costs necessary to provide the element at least cost, given the existing network configuration. NTIA does not, however, favor a green field approach to interconnection pricing, as that will understate the ILECs' true costs. To the extent that the "forward-looking" component to Justice's TSLRIC contemplates a green field approach, NTIA urges that it be used as a lower bound to structure negotiations between LECs and interconnectors. NTIA and Justice agree that interconnection rates must reflect joint and common costs. See id. at 27 ("TSLRIC rates may need to be adjusted to permit recovery of forward-looking and common costs that may not be included in the sum of element-by-element TSLRIC rates"). In sum, if Justice is, in fact, not suggesting a green field approach to interconnection pricing, our approaches are quite similar.
 The service in question here is, of course, telephony service. Consequently, the derivation of an ILEC's LRAIC must not include investments "in facilities 'over-designed' to enable the present or future delivery of cable television or other video services, 'official networks' overdesigned to facilitate the potential future provision of interexchange services, and other investments designed to meet the expected incremental demands or technical requirements of non-basic or enhanced services." Comments of AT&T Corp. at 59
 In the long run, all costs, including capital costs, are variable. Long run costs include a "fair" return on invested capital. This return is correctly thought of as a cost; if it were not paid, those investment dollars would migrate to other opportunities that did pay a fair return. Long run costs must, therefore, reflect the opportunity cost of capital. This also includes the ILEC's costs of unbundling.
The reference to incremental cost leaves open the "size" of the increment. To the extent that average costs vary with the quantity of service produced, the increment should be each additional unit of service provided (i.e., each additional customer). In practice, however, if may be difficult, if not impossible, to measure long run average cost at such a fine level. Instead, the regulator will likely treat the entire service or functionality (all subscribers served) as the increment, and estimate average costs by dividing the total long run costs associated with the service (or functionality) by the number of subscribers.
 To the extent that State regulators are certain that an ILEC has little incentive to reach interconnection agreement or has too much bargaining power over potential interconnectors, the threat of TSLRIC pricing may be an effective motivational tool. We prefer reliance on voluntary negotiation to establish interconnections terms. As discussed below, in many cases, the Act contains sufficient incentives for both interconnectors and ILECs to reach agreement. We also recognize, however, that State regulators may need latitude to determine interconnection rates in those instances when voluntary negotiations fail. See, e.g., Washington Util. & Transp. Comm'n v. US West Comm., Docket No. UT-950200, Fifteenth Supp. Order 89 (Apr. 11, 1996).
 Here we allude to TSLRIC estimates that do not already include an "overhead factor" to account for shared costs. For Hatfield's most recent estimates, the regulator would subtract the six percent overhead factor to recover "pure" TSLRIC estimates. ILECs would be required to justify costs in excess of these lower estimates.
 As the Justice Department points out, TSLRIC should be adjusted upward to account for forward-looking joint and common costs, whenever such costs exist. DOJ Comments at 27, 31-32. NTIA believes that the Commission should determine what permissible categories of such joint and common costs are.
 See Bridger Mitchell, Werner Neu, Karl-Heinz Neumann, and Ingo Vogelsang, "The Regulation of Pricing of Interconnection Services" at 116 (Mitchell et al.) (on file with NTIA).
 The same would be true of a rate fixed by the State via arbitration.
 The RA [Regulatory Authority] should indicate that, if no agreement was reached and it were asked to make an ex-post determination, it would determine a charge in that range on the basis of its assessment of the demand conditions in that market. This would provide proper incentives for the two sides in the negotiations. Not knowing what the RA would do in case of failure of negotiations, both would have a preference to reach a settlement on their own accord. Of course, if either party speculated that it would have a good chance that its view on charges would be confirmed by the RA, it might opt to let negotiations fail and rely on the RA's decision. This would have to be accepted as a legitimate part of the process.
Mitchell et al., supra note 52, at 114.
 NTIA believes that the prospect of interLATA entry will give the BOCs, at least, incentives to negotiate in good faith. We recognize, nevertheless, that the BOCs' incentives in this respect may decline after they have been certified to provide interLATA service. Interconnectors may be protected, to some extent, by the BOCs' existing agreements in that market. To the extent, however, that those agreements are unsuitable for the late entrant, the State regulator may have to choose rates in response to failed negotiations. In extreme circumstances, we see no reason why the Commission could not revoke an ILEC's certification to provide interLATA service.
In general, we favor agreements that give ILECs no incentive to alter their behavior after receiving interLATA certification. Self-executing contracts, which spell out penalties for readily observable ILEC violations with regard to service quality, may be helpful. If these are among the first contracts signed by ILECs, the self-executing provisions may also afford protection for subsequent entrants.
We also strongly recommend that the Commission include as part of the public interest standard governing BOC interLATA entry under Section 271 a requirement that the BOCs have in place processes and procedures for on-going interconnection negotiations, that they have established a practice of bargaining in good faith, that their agreements with interconnectors be self-enforcing, and that their business dealings pursuant to those agreements be found to be in good faith.
 Once a price for an interconnection arrangement or unbundled network element has been established, whether by negotiation or arbitration, that rate ought to become the ceiling of the zone of reasonableness in subsequent negotiations between the same parties about the same arrangement or element. In other words, in any negotiation triggered by the expiration of an interconnection agreement, the zone of reasonableness for the rates for any arrangement or network element that was part of the old agreement will be defined by TSLRIC and the applicable rate in the expired agreement. It would be reasonable, however, to permit some flexibility in that upper bound to reflect changes in ILEC costs over time. One solution, at least for price cap regulated ILECs, would be to allow adjustments to the upper bound in accordance to the price cap escalator formula applicable to the ILEC involved. The parties would remain free, of course, to negotiate (and a State commission could establish) a new rate within the new zone of reasonableness.
 This latter option would minimize the chance of a mistaken choice in those instances where, for example, one party's price lies at one extreme of the zone and the other party's offer falls just outside the other extreme.
 See Notice 268 (requesting comment on "final offer" arbitration).
 A similar approach is suggested by Mitchell et al.:
The percentage mark-ups on top of LRAIC for interconnection services should vary between zero, as the lower limit, and, as the upper limit, the minimum uniform mark-up, that is, that common mark-up that, when applied to the LRAIC of each service, would lead to revenues that cover all costs, including common costs, and all other revenue requirements.
Id. at 113.
 One could, conceivably, compute the ILEC's total costs before and after it provided a service or functionality to an interconnector to determine that entrant's individual TSLRIC. The same interconnector, however, might face different TSLRICs depending on when it entered the market. That is, the first entrant might face a relatively large TSLRIC, as the ILEC reconfigured its network to accommodate interconnection while subsequent entrants might bear much lower burdens. We think it best, therefore, that the ILEC compute a comprehensive TSLRIC for a service, and then apportion that cost among users.
 Mitchell et. al., supra note 52, at 106.
 Gerald Brock, "Interconnection and Mutual Compensation With Partial Competition," in The Economics of Interconnection 14 (Apr. 1995) (prepared for Teleport Communications Group).
 More generally, Brock notes that:
If the established price for a channel of given capacity is near the real cost, then the NYNEX-Teleport arrangement provides an attractive model for general interconnection issues. It would approach a cost-based interconnection fee for both peak and off peak traffic, leading to economic efficiency and opportunities for pricing innovations.
Id. at 15.