Full competition in telephony requires the solution of numerous problems, some of which appear truly intractable. This paper deals with three of the most serious difficulties that competition in telephony must overcome: interconnection among local exchange service providers, interconnection between local exchange and interexchange service providers, and the transition from monopoly to competition. This paper shows that customer ownership of the local loop provides an elegant, uniform solution to these problems; it causes each of these problems to virtually disappear. The paper also discusses the likelihood that customer ownership of the local loop will create a competitive loop construction industry. This does not directly solve an obvious problem with competition in telephony, but it is clearly desirable, because it permits loop construction to be demand driven rather than supply driven. The paper concludes with an explanation that customer ownership of the local loop does more than solve several individual problems with competition in the provision of telephony. Rather, it constitutes an elegant, integrated solution to many of the problems with competition in telephony.
In order for competition in local telephony to be successful, multiple local exchange companies (LECs) must interconnect with each other, to form a network that appears seamless to customers. Interconnection among competing providers of local telephony, however, has become a nearly intractable problem, because the interconnecting parties are typically of two very different types. On the one hand, there is an incumbent who owns a complete infrastructure. On the other hand, there are competitive entrants, who wish to provide service in competition with the incumbent, but who are forced to use at least part of the incumbent's infrastructure--namely, the local loops--because constructing a complete infrastructure is prohibitively expensive. The situation, thus, is as follows: an entrant constructs a limited infrastructure--perhaps in the downtown business area--consisting of a switching center ("central office") and local loops to its customers, and then seeks to interconnect with the incumbent. Interconnection is required for its customers to place calls to, and receive calls from, everyone's customers. The interconnection that the entrant seeks, however, appears especially insidious to the incumbent: the entrant wishes to, in effect, "rent" the local loops of the incumbent's customers--in order to compete with the incumbent!
It is, therefore, not surprising that it is rare for an incumbent and a competitive entrant to agree on interconnection terms: It is in the incumbent's self-interest to refuse to interconnect at all, or to charge such a high rent that the entrant cannot successfully compete. It is in the entrant's self-interest to obtain interconnection for nothing, or, at least, for less than the cost of building and maintaining its own local loops. No consensus exists concerning the proper charge for interconnection, yet the price of interconnection is critical to the success of competition. If the incumbent charges the "proper" price for interconnection, then entrants that are more efficient than the incumbent will have incentive to construct their own network (because by doing so they will have lower costs), and provide facilities-based competition. If interconnection is priced too low, however, then the incumbent will not recover its costs of loop construction and maintenance, and an inefficient entrant will have incentive to resell the incumbent's network services rather than construct its own network. If interconnection is priced too high, an entrant will be unable to price its service at a level that allows it to compete with the incumbent.
This is not a problem that has a simple regulatory solution. The problem is not that the proper price for interconnection is well known, but that neither the incumbent nor the entrants are willing to accept a regulatory mandate that the entrant pay that price. Rather, the problem is that no one knows what that proper price is, nor does anyone know how to determine it.[1] Indeed, analysis of the assumptions underlying the need for interconnection leads one to conclude that pricing interconnection is a dilemma: one can either be unfair to the incumbent or unfair to the entrant, but not, apparently, simultaneously fair to both.[2]
There is a straightforward solution to the dilemma of interconnection pricing: customers should own their local loops. If customers own their local loops, it is clearly they alone who have the right to decide which LEC should originate and terminate their traffic, on their local loop. Interconnection among competing local service providers would still be necessary, of course. However, it should not be difficult for companies to negotiate, or for regulators to mandate, its terms and conditions. For, when customers own their local loops, service providers negotiate from relatively symmetrical positions: each provides service to its customers, but none is required to rent the use of its competitor's bottleneck facility--from its competitor--at a price that includes its competitor's monopoly rent. Or, none is required to rent the use of its bottleneck facility to a competitor, at a price that does not cover the large embedded cost of constructing that bottleneck facility.
In more detail, if customers own their local loops, service providers will probably need to interconnect at two places: the frame, and the trunk side of the switch. In the former case, the "interconnection" is merely the running of wires from the positions on the frame where LEC A's customers' local loops terminate to LEC A's switch, and the running of wires from the positions on the frame where LEC B's customers' local loops terminate to LEC B's switch. The additional costs due to this "interconnection" are minimal; the frame becomes a jointly-owned piece of hardware.[3]
Interconnection will be desirable at the trunk side of the switch because LEC A's customers will, presumably, want to place calls to LEC B's customers (and, of course, vice versa); the former are served by LEC A's switch, and the latter are served by LEC B's switch. In this case, there is a non-zero additional cost for interconnection, but it is not high: it is no more than the cost of a LEC adding an additional central office to its (local) network. Here, again, it seems reasonable to share the cost of interconnection among the LECs involved.[4]
The issue of paying for the cost of terminating calls on a competitor's network should disappear, because that cost--once the construction and maintenance of the local loop is removed--is virtually zero. Or, even if it is not zero, it is probably considerably less than the cost of billing such call termination. And, when customers own their local loops, incumbents have little justification for the argument that they are providing more value to their competitors than they receive from them, because it is the customers themselves, rather than the LECs, that are "providing" the local loops for call termination.
The preceding section has shown that--without customer ownership of local loops--interconnection among LECs is especially problematic, because the incumbent is being asked to provide an input to the entrant (the local loop), in order to enable the entrant to compete with the incumbent. When LECs interconnect with interexchange carriers (IXCs), however, no such problem exists, because LECs and IXCs provide complementary services. In fact, interexchange service would be useless without local service, and local service would be reduced in value, at least, if it were provided without connection to interexchange service providers. Nevertheless, interconnection between LECs and IXCs is not without problems. This section describes the problems with interconnection between LECs and IXCs, and shows that customer ownership of the local loop reduces the severity of those problems.
Until 1930, long distance (interexchange) service in the U.S. was billed using "board to board" accounting. Under this accounting method, the interexchange carrier received 100% of the revenue from interexchange service. The reasoning for board to board accounting was as follows: the provision of a connection from a customer's premises to the central office, and the provision of a connection from the central office to a customer's premises, are the normal functions of a local exchange company. The fact that those connections are sometimes for the origination and termination of interexchange service is irrelevant: the LEC is paid to provide connections between customers' premises and central offices, so it should do so. The LEC plays no part in carrying traffic between central offices, and so receives none of the funds for those services.
State regulators realized that the physical plant necessary for local service--including telephones, inside wiring and local loops (which were, at the time, still owned by the LECs)--was also necessary for, and also used in the provision of, long distance service. They argued, therefore, that IXCs should pay a part of the cost of the "local" physical plant.[5] In 1930, the U.S. Supreme Court ruled in favor of this argument: because the local network is used jointly for local and interexchange traffic, its cost must be paid jointly by LECs and IXCs.[6] This decision mandated the use of "station to station" accounting for long distance service: the cost of such service is not merely the cost of carrying traffic from the originating central office's switch ("board") to that of the terminating central office. Rather, the cost of long distance service must include the cost of carrying traffic all the way from the originating telephone ("station") to the terminating one; the process of determining how much of the cost should be born by LECs and how much by IXCs is called separations. This issue involved an interesting trade-off between cost and authority: if the states had been willing to retain board to board cost allocation, then there would have been no role for federal regulators in setting rates of local telephone companies. However, by arguing for station to station allocation, state regulators ceded some of their control to federal regulators--presumably because they believed that lower local rates were more desirable than maintaining complete control over those rates.[7]
Unfortunately, separations are not easily specified. The earliest separations case was in New York State, in 1910. The New York Public Service Commission concluded,
It is impossible to determine the cost of the toll service separately from that of the local service for the reason that the greater part of the cost of both is joint cost and there is no way of allocating the proper portion of the joint cost to each branch of service.[8]
Despite this finding in New York, separations have been ubiquitous at least since Smith vs. Illinois. Before 1984, when the majority of local traffic and nearly all long distance traffic was carried by AT&T and its affiliates, much of the impact of separations was on which regulatory body--state or federal--had jurisdiction over, rather than on the economic welfare of, the telephone company.[9] After divestiture, however, separations represented real money, paid by IXCs to LECs. These payments currently are in the form of access charges, which are approximately five cents per minute each, to the originating and terminating LECs.
Part of this fee makes sense, and is relatively straightforward to determine: the part that is usage sensitive. This includes switch sizing, and some parts of maintenance, buildings, etc. However, the major cost of local service--the local loop--is wholly traffic insensitive: the entire cost of the local loop is in its construction and maintenance--even if it is never used! In other words, the wire does not wear out faster if it carries more electrons. Thus, usage-based separation of costs between local and interexchange traffic on the local loops is impossible to justify quantitatively. As telephone companies upgrade their networks, this problem gets worse: telephone companies often contemplate (at least) an upgrade that will add the capacity for carrying high-speed data (including video), as well as voice traffic, over their local loops. In these cases, the issue of assigning costs to different services reappears, with no obvious solution in sight:
There is perhaps no better example of the inability to learn from history than the current disputes over the proper apportionment of fixed and common costs. We appear to be preparing to replay the futile search in the 1960s and 1970s for methods to apportion AT&T's fixed and common costs, an exercise that ended in failure. It is essential that alternative regulatory schemes be developed to avoid this fruitless exercise.[10]
If, however, customers own their local loops, this becomes a moot point: customers would purchase and maintain (or, hire someone else to maintain) their local loops--the amount of use of which is completely unrelated to the costs of construction and maintenance. Thus, the determination of the payments that IXCs, or other service providers, should make to LECs for originating and terminating their traffic would be relatively straightforward. It would require only determining the total cost of the local plant and the total number of minutes that it is used, in order to calculate a per-minute usage charge for any user--IXC, LEC, video provider, etc. The problem of allocating a non-usage sensitive cost--the local loop--among several users is, thus, completely avoided.
In 1974, the U.S. Department of Justice (DOJ) filed an antitrust suit against AT&T, on the basis of claims that AT&T had aggressively resisted interconnection with competitors in long distance service. The theory of the DOJ's case reflected their belief that long distance service is similar to customer premises equipment (CPE), which by then was no longer a part of the regulated telephone monopoly.[11] According to this model, customers should be free to purchase any long distance service they wish, and connect it to their network--period. Implementation of the CPE model for interexchange service, however, would have eliminated the substantial amount of money--the exact amount of which was determined by the separations process--that was transferred from long distance to local service providers under then current agreements. The elimination of this money transfer would have increased local telephone rates substantially, and so the CPE model was opposed by state regulators.[12]
Without something like the CPE model of long distance service, we are left with a complex system of separations, in which it is necessary to assign the various joint and common costs of local and long distance service (as well as of any new services, such as video on demand, or home shopping, or cable TV) to the respective services. It is extraordinarily difficult to assign costs proportionally to different services, and to re-assign those costs when new services are provided over existing infrastructure. It is even more difficult to do such a cost assignment in light of the fact that the majority of the joint and common cost is the local loop, which is not traffic sensitive. However, if customers own their local loops, then the CPE model is straightforward to implement: the majority of the joint and common costs are paid for by each customer who purchases a local loop, and remaining joint and common costs are either small enough to be probably negligible, or traffic sensitive so that they can be easily apportioned on the basis of use, or both.
Once it is accepted that competition in telephony is desirable, it is essential to devise a plan for the transition from the current state, where service is provided by regulated monopolies, to a future in which service is provided by competing firms. The reason such a plan is essential is that competition will not be starting from a neutral position. So long as the incumbent monopolist is permitted to be a player in the competitive future, the early competitive period will include competitors with vastly different amounts of market share, name recognition, expertise and experience. This section discusses some problems of the transition to competition, and argues that customer ownership of the local loop can play a key role in their solution.
Currently, local service prices are only vaguely based upon costs. Thus, while prices may be identical within an entire metropolitan area, costs most likely are not: it is virtually always more expensive to provide service to an outlying suburban neighborhood than to a central city neighborhood, or a downtown business district. In addition, business prices are often about twice as high as residential prices, but the cost of providing service to businesses is probably similar to, or modestly greater than, the cost of providing residential service.[13]
The reason cost-based pricing (or the absence thereof) is an issue for the transition from monopoly to competition is that competitive entrants will, clearly, choose to enter exactly those markets where prices are above costs, and to ignore those markets where prices are below costs. This is likely to result in market failure, for two reasons. When the incumbent's price is above its cost, an entrant need not be more economically efficient than the incumbent in order to capture market share--it need only be sufficiently efficient to produce at a cost below the incumbent's price (this is often called "cream skimming," and is done by entrants). When the incumbent's price is below its cost, an entrant can be more economically efficient than the incumbent but still refrain from entry: it needs to be sufficiently efficient to have costs below the incumbent's price, not just below its cost (this is often called "predatory pricing," and is done by the incumbent).
So, if telephone service were priced at cost, there would be little worry of inefficient entry or of efficient non-entry. Since the major difference in the cost of providing local service is the cost of the local loop, customer ownership of the local loop would virtually eliminate the need for price differentials, thereby virtually eliminating the dangers of cream skimming and predatory pricing. As mentioned above, the differences in time-of-day, and volume, usage patterns between business and residential customers--or even among groups of business or residential customers--may still require pricing differentials. These differentials, however, would be usage-based, and so could be easily applied to anyone--business or residential customer--whose usage generates demands for additional busy hour capacity. In New Zealand, for example, Telecom Corporation of New Zealand, charges 3.55 cents per minute for calls during peak hours, and 0.44 cents per minute for calls during nights, weekends and holidays.[14]
There are two major reasons for the absence of cost-based pricing in local telephone service. First, there is political pressure on state regulators (who are typically appointed by elected state officials) to maintain low residential telephone rates. Second, non-cost-based pricing is used to fund universal service--the availability of telephone service to all, at a reasonable price. The current industry/regulatory environment achieves a "reasonable" (below cost) price for high cost customers--primarily those who are far from the central office and have long local loops--by charging an above cost price to low cost customers. This is the reason that cream skimming has deleterious effects on universal service, as currently implemented. Namely, cream skimming removes from the incumbent's market the high price, low cost customers (for an entrant need not be particularly efficient to serve low cost customers at a price lower than the incumbent's), the profits from whom are used to subsidize low price, high cost customers. Thus, with the advent of competition, it is likely that competitive entrants will choose to provide service primarily to low cost, high revenue customers--leaving the incumbent with no customers paying above-cost prices whose profit it can use to subsidize its customers paying below-cost prices.
So, there are two types of problems posed by non-cost-based pricing for the transition to competition. First, inefficient entry may be encouraged and efficient entry may be discouraged; second, subsidies for high cost users will be unlikely to survive. Customer ownership of the local loop eliminates both of these problems by eliminating the need for, or the temptation of, non-cost-based pricing. A call from a business customer is no more expensive to switch than is one from a residential customer; likewise, a call from a remote area is no more expensive to switch than is one from a central city. It is the costs of the local loops that differ for different customers, in proportion to the lengths of those loops. If customers own their local loops, the cost of providing service to all customers will be approximately the same. Then, there will be no low cost and high cost customers, so there will be no need for the former to subsidize the latter, and there will be no danger of discouraging efficient entry and encouraging inefficient entry.[15]
Even if prices were based on cost, however, the transition to competition--in the current telecommunication environment--would not be simple. For, in order for an entrant to compete, it must make large investments in infrastructure.[16] In New Zealand--a country of about 270,000 square km and about 3.5 million people[17]-- BellSouth's Director of Strategic Planning has recently estimated that the cost of entry into the telecommunication market is at least $NZ 250 million.[18] In the United States, about 60% of the book value of the LECs' plant is in transmission equipment and wire and cable, "most of which represents local loop plant."[19] Clearly, eliminating the cost of constructing local loops would significantly decrease the cost of entry into the telecommunication market. And, equally clearly, if customers owned their local loops, the cost of constructing local loops would be eliminated from the cost of entry into telecommunication.
Another difficulty with the transition to competition--one which is also related to the cost of duplicating the local loop--is the issue of unbundling and resale. Typically, part of the legislative underpinnings of competition require the incumbent to separate various elements of its retail service package, and to resell them--individually--to competitive entrants. For example, in the United States, the Telecommunications Act of 1996 specifies that the incumbent has
(3) The duty to provide, . . ., nondiscriminatory access to network elements on an unbundled basis at any technically feasible point . . . . An incumbent local exchange carrier shall provide such unbundled network elements in a manner that allows requesting carriers to combine such elements in order to provide such telecommunications service.(4) The duty to offer for resale at wholesale rates any telecommunications service that the carrier provides at retail to subscribers who are not telecommunications carriers. . . .[20]
The justification for a resale requirement is that it reduces the cost of entry: a competitive entrant can "lease" network elements that it is unable (or unwilling) to provide itself, and package those elements with the ones that it does produce itself, to create a complete service package. The justification for an unbundling requirement is that it permits the entrant to lease only those network elements that it does not produce itself, and prohibits the incumbent from determining what combination of elements an entrant will be able to resell.
There is a major difficulty with resale, however: it provides only modest competition, at best. For, if major portions of an entrant's retail service package are merely repackaged and resold elements of the incumbent's products, then what is the likelihood that the entrant will be more efficient than the incumbent? Or, a major purpose of competition is to increase economic efficiency; however, an entrant who competes primarily by reselling the incumbent's service elements will be unlikely to provide competitive pressure on the incumbent. In fact, unbundling and resale has no built-in mechanism for distinguishing between inefficient entrants, that succeed only because of the unbundling and resale requirements, and efficient entrants, that are protected from unfair competition by unbundling and resale requirements.
It is possible to create only temporary unbundling and resale requirements, with the understanding that an entrant is to be provided with competitive assistance only to help it become viable; after a certain period of time--or after attaining a certain market share--such requirements would be removed, and the entrant would have to compete on equal terms with the incumbent. In the U.K., for example, Mercury's interconnection price rises sharply once its payments to British Telecom exceed 7% of BT's "corresponding aggregate revenues."[21] This, however, introduces a new problem: exactly when should the temporary period end? There is no clear, principled method of determining how long an entrant should receive favorable treatment; it is clear that the entrant will argue forcefully for such treatment to last as long as possible, while the incumbent will argue--equally forcefully--for it to last as short as possible. In this case, the success of an entrant is likely to rely as much (or more) on its legal strategy as on its economic efficiency.
Two other problems with unbundling and resale requirements are related to the symmetry of their application, and what kinds of incentives they provide. The symmetry issue is, should unbundling and resale requirements pertain to entrants as well as to incumbents? Entrants, of course, will say "no" (as does the U.S. Telecommunications Act of 1996): they may have built fiber optic rings in the business centers of large cities, and they see these rings (probably correctly) as their major competitive advantage. They will probably couch their objections in terms of economic efficiency: if they are forced to resell their infrastructure to the incumbent, what incentive will the incumbent have to create its own modern infrastructure?[22] However, this objection is equally applicable to the incumbent: if the incumbent is forced to resell its infrastructure to an entrant, what incentive will an entrant have to create its own infrastructure? Enacting asymmetrical regulation is extraordinarily problematic; it virtually always results in ad hoc application of its details. Such application, even when done by people of intelligence and good will, is rarely as defensible as symmetrical regulation. With regard to regulation of telecommunication in general,
. . . symmetric regulation should be adopted for the increasingly competitive telecommunications sector. This is required to provide market-based price signals which induce efficient investment and entry. All forms of asymmetric regulation contain an intrinsic bias toward some firms or technologies and run the risk of imposing large productive efficiency costs.[23]
With customer ownership of the local loop, however, the issue of unbundling and resale is at least greatly mitigated. For when customers own their local loops, a potential competitor is relieved of a large amount of the cost burden of entry. In fact, it is probably not implausible to claim that an entrant who cannot afford to raise the capital for switches, inter-office trunks and their maintenance is unlikely to succeed, and that providing it with favorable treatment--for any period of time--is likely only to encourage inefficient entry.
If customers own their local loops, then it is reasonable to assume that loop construction will become a competitive industry. Thus, just as there are multiple providers of television sets, or high fidelity audio components, there can be multiple providers of local loops. This has two major advantages over the current situation. First, there are the usual benefits of competition: efficient use of resources, cost-based pricing, technological innovation, and consumer choice. The second advantage is perhaps more subtle: it makes loop construction demand-driven rather than supply-driven. In other words, customers will--if the loop construction industry is truly competitive--be able to choose the technology they prefer for their local loops. This is an advantage for producers as well as for consumers: it is an advantage for consumers because they will be able to arrange for the loop technology that meets their needs or desires; it is an advantage for producers because they will not have to gamble on implementing a particular loop technology, and hope that customers will be willing to pay for it by subscribing to services that will pay for it. With a competitive loop construction industry, it is not unreasonable to expect that some customers will choose "ordinary" copper twisted pair, while others will choose coaxial cable, fiber optic cable, or radio. So long as customers pay for their choices, providers have no reason to construct only a single type of local loop.
A major strength of customer ownership of the local loop is that it provides an integrated solution to the problems of competition in telephony. Thus, it may be possible to solve the problem of LEC-IXC interconnection, for example, without customer ownership of the local loop; or, it may be possible to solve the problem of competitive entry into telephony without customer ownership of the local loop. However, without customer ownership of the local loop, each problem requires its own, carefully tailored solution, and the solution to one problem would not necessarily have any relationship--symbiotic or otherwise--with the solution to another problem. Customer ownership of the local loop, on the other hand, is a remarkably simple, single solution to all of these problems. The remainder of this section briefly describes how the various problems interact, and reminds the reader that customer ownership of the local loop solves each one.
Ideally, competitive entry into an industry is a non-issue: if an entrant thinks that it can produce a better product than is currently available, or that it can produce a similar product at a lower price, then the entrant tests its belief by entering the market. In telecommunication, however, there is the combination of a history of government protected monopoly and a capital-intensive industry, so the incumbent at least appears to have a significant competitive advantage. This advantage appears so significant that potential entrants are virtually unanimous in claiming that more than simple legislation permitting competition is required for the entrants and the incumbent to compete on a relatively equal basis.[24]
It is probably theoretically possible for a government to devise a mix of rules and regulations that results in approximate parity between the incumbent and competitive entrants. However, such a set of rules would be extraordinarily complex, would have to be tailored carefully to each incumbent and each entrant, and would have to change over time, as the incumbent lost the competitive advantage that was due merely to its former monopoly. On the other hand, customer ownership of the local loop removes such a large part of an entrant's financial burden that it is at least arguable that the incumbent has no significant competitive advantage beyond its name recognition and reputation--which, in some cases, hardly constitute a competitive advantage. Thus, competitive parity would be created automatically, with no need for the initiation, examination, or continuation of measures whose sole purpose would be to compensate for past monopoly.
The issue of unbundling and resale is related to the transition from monopoly to competition. The argument for resale and unbundling is that an entrant is unlikely to be able to afford to construct an entire infrastructure prior to providing any service, so incumbents are required to unbundle their network elements, and permit entrants to resell only those elements that the entrants are unable to provide themselves. The argument against resale and unbundling is that competition based upon resale of an incumbent's network is really no competition at all. For such competition is based upon the efficiency of the incumbent, at least for those portions of the network that an entrant resells. And it is not clear that society benefits from the entrance of a competitor that is more efficient than the incumbent in only one aspect of service provision, and that depends upon the incumbent for the remainder of its service elements.
As with competitive entry, it is, in principle, possible to solve all the problems with resale and unbundling. For example, it is probably possible to decide whether all competitors should unbundle and resell service elements or just the incumbent; it is probably possible to decide just how long resale should be available--perfectly balancing the heavy initial burden upon competitive entrants against the greater efficiency of facilities-based competition; it is probably possible to decide just which service elements should be resold, and which it is reasonable to expect an entrant to provide upon entry; etc. However, it is certainly simpler have customers own their local loops.[25] In this case, no decisions concerning the amount, symmetry, or duration of unbundling and resale need be made. For, if customers own their local loops, the cost to entrants is sufficiently reduced--to about 40% of the cost without customer ownership of the local loop[26]--that entrants can be expected to provide the remainder of the network and service elements upon entry.
The major difficulty of interconnection among LECs is closely related to the problem of unbundling and resale. This difficulty is that the complete infrastructure required to deliver service to the customers premises--primarily the local loop--is so expensive to build and so unnecessary to duplicate, that entrants are virtually always reduced to negotiating interconnection agreements whereby they "rent" the local loops of the incumbent. This leads directly to the difficult problem of what price entrants should pay for the use of these local loops. Typically, incumbents think that the price should be high, and should include their lost opportunity cost, while entrants think that the price should be low, for costs are low, and incumbents as well as entrants derive benefit from interconnection.
It may be theoretically possible to discover a price for interconnection such that efficient entry is encouraged and inefficient entry is discouraged. However, experience with attempts at interconnection pricing suggest that it is difficult or impossible to reach mutually satisfactory agreements on this issue; it is likely to be considerably simpler for customers to own their local loops. In this case, interconnection among competing service providers would still be required, and the appropriate manner of sharing the costs of that interconnection would still have to be determined, but entrants would no longer be using local loops that belong to their competitors, in order to compete with those competitors. Therefore, the issue of lost opportunity--which is the major unresolved issue in pricing interconnection--would disappear.
The major problem with interconnection between LECs and IXCs is how to divide joint and common costs among the services. It is clear that intercity trunks, for example, are used only for long distance service, but that interoffice (intracity) trunks may be used for both local and long distance services. Historically, joint and common costs have been divided among services on the basis of relative minutes of use. This is not a cost-based solution, however, because the majority of the joint and common costs of local and long distance service are related to construction and maintenance of local loops--costs which are not usage sensitive. The problem is further exacerbated by the desire to deliver new services over existing infrastructure; each time a new service--video, for example--is provided over existing infrastructure, a new allocation of joint and common costs will be required.
It is possible to allocate joint and common usage sensitive costs among services--based upon relative use, presumably. However, it is not even theoretically possible to devise a sensible strategy for allocation of non usage sensitive joint and common costs among various services. Clearly, customer ownership of the local loop is far more straightforward than any other proposed solution to this problem. When customers own their local loops, allocation of joint and common costs among various services is moot: customers purchase a local loop because they desire service. They purchase only those services which they desire. In addition, if customers desire to purchase only certain services, they can design their local loop for those services alone, and purchase a loop which is sufficient to carry those services--local and long distance, for example--and not others--video on demand, or Internet connectivity, for example. 7. Summary and Conclusions This paper has discussed three major problems with competition in telephony: interconnection among local exchange service providers, interconnection between local exchange and interexchange service providers, and the transition from monopoly to competition. It may be possible to solve each of these major problems without customer ownership of the local loop. However, such a set of solutions would probably be ad hoc, difficult to specify and administer, and complex. On the other hand, it seems that all of these problems are related to the high cost of local loop construction. Thus, it is not surprising that customer ownership of the local loop provides a straightforward, integrated solution to them. Customer ownership of the local loop is also likely to result in a competitive, demand driven, loop construction industry.
Alleman, J.H. The Pricing of Local Telephone Service. U.S. Dept. of Commerce, Office of Telecommunications, OT Special Publication 77-14, 1979.
Albon, R. "Interconnection Pricing: An Analysis of the Efficient Component Pricing Rule." Telecommunications Policy 18(5), 1994, pp. 414-420.
Arellano, M. "Exploiting the LECs' Achilles' Heel." Telecom Strategy Letter, July, 1995, p. 83.
Baseman, K.C. & Van Gieson, H. "Depreciation Policy in the Telecommunications Industry: Implications for Cost Recovery by the Local Exchange Carriers." MiCRA, Inc., December, 1995.
Baumol, W.J. & Sidak, J.G. Toward Competition in Local Telephony. Cambridge, MA: The MIT Press, 1994.
Brock, G.W. Telecommunication Policy for the Information Age: From Monopoly to Competition. Cambridge: Harvard University Press, 1994.
Brooks, J. Telephone: The First Hundred Years. New York: Harper & Row, Publishers, 1976.
Crandall, R.W. & Waverman, L. Talk is Cheap: The Promise of Regulatory Reform in North American Telecommunications. Washington, D.C.: The Brookings Institution, 1995.
Crook, J. "Competition and Interconnection: Successes and Challenges. A Practical Perspective." Paper presented at ITS Workshop on Interconnection, Wellington, N.Z., March, 1995.
Economides, N. and White, L.J. "Access and Interconnection Pricing: How Efficient is the 'Efficient Component Pricing Rule'?" Antitrust Bulletin, 40(3), 1995, pp. 557-579.
Epstein, R.A. Simple Rules for a Complex World. Cambridge, MA: Harvard University Press, 1995.
Epstein, R.A. Takings: Private Property and the Power of Eminent Domain. Cambridge, MA: Harvard University Press, 1985.
Fischer, C. S. America Calling: A Social History of the Telephone to 1940. Berkeley: University of California Press, 1992.
Gabel, R. Development of Separations Principles in the Telephone Industry. Michigan: Michigan State University, 1967.
Hatfield Associates, Inc. "The Cost of Basic Universal Service." July, 1994.
Kahn, A.E. & Taylor, W.E. "The Pricing of Inputs Sold to Competitors: A Comment." The Yale Journal on Regulation, 11(1), 1994, pp. 225-240.
MacMeal, H.B. The Story of Independent Telephony. Chicago: John F. Cuneo Company, 1934.
Schankerman, M. "Symmetric Regulation for Competitive Telecommunications." Information Economics and Policy 8:1(1996), pp. 3-23.
Schechter, P.B. "Telecommunication in New Zealand: Competition, Contestability and Interconnection." Submitted to Pacific Telecommunications Review.
Stehman, J.W. The Financial History of The American Telephone and Telegraph Company. New York: Augustus M. Kelley, 1967.
United States Central Intelligence Agency. The World Factbook 1994.
U.S. Bureau of the Census. Statistical Abstract of the United States: 1995 (115th edition). Washington, D.C., 1995.
Walker, D. & Solomon, J. "The Interconnection Imperative: E pluribus unum." Telecommunications Policy, 17(4), 1993, pp. 257-280.