This paper was presented at the Columbia University CITI seminar on "The Role of Wireless Communications in Delivering Universal Service", October 30, 1996. © The Telmarc Group, Inc, 1996, All Rights Reserved. The Telmarc Group, Inc., 24 Wo
odbine Rd, Florham Park, NJ 07932,
mcgarty@rpcp.mit.edu
The Telmarc Group, Inc.
Abstract
Wireless local loop is a vision of how wireless technology may be deployed in the provisioning of local telephone service. Both the large existing entities such as AT&T and the entrepreneurial entrants such as NextWave and others have look at taking t he PCS spectrum and using the new technologies to allow for the provision of full local loop capabilities in a wireless fashion. This paper discusses the economic viability of such a plan.
There currently are several different strategies aimed at the presentation of local telecommunications. One is the strategy of attempting to go head to head with the existing LEC on a local loop only basis. The other is the bundling of this with other se rvices, such as long distance and television delivery to attempt to obtain scale economies in the delivery of the services. The latter approach is clearly expressed by both AT&T and MCI.
The key issue is initially one of cost. Can wireless be delivered at a lowered costs per unit service than can the existing wire based local loop. Previous analyses have shown that wire base systems have capital per subscriber as high as $1,800. Wireless is at about $250 which does not include the handset. However the incumbent has these as sunk costs whereas the new entrant has these as entry costs. This is a significant difference. Add to this the license costs which in many cases equals the cost of ca pital plant.
There is also the issue of bundling. PCS generally is a voice only system, although some data is available. LMDS is a voice, video and data system at 28 GHz. The question is, does the ability to bundle significantly increase the entry opportunity.
Finally, all of these wireless financial numbers are exclusive of any Universal Service coverage commitment. If one adds a surcharge on all players then the approach is simple and each customer is taxed for the Universal Service Fund. If however there is a requirement to physically cover, then this is generally uneconomic. We shall show the value per PoP versus population density. It is clear that in most wireless systems at densities at less than 100 PoPs per square mile there is no economic value to th e business. The policy issue is related to providing Universal Service but not forcing Universal Coverage by all participants.
2. Universal Service
Universal Services is the mandate to provide services by any carrier to any person not individually financially able to obtain the service in the area in which the inhabit. Namely the low income and rural customers. The universal services provisions are a s follows:
" (b) UNIVERSAL SERVICE PRINCIPLES- The Joint Board and the Commission shall base policies for the preservation and advancement of universal service on the following principles: (1) QUALITY AND RATES..... -(2) ACCESS TO ADVANCED SERVICES - ..... (3) ACCESS IN RURAL AND HIGH COST AREAS - ...... (4) EQUITABLE AND NONDISCRIMINATORY CONTRIBUTIONS ......
(c) DEFINITION (1) IN GENERAL- Universal service is an evolving level of telecommunications services that the Commission shall establish periodically under this section, taking into account advances in telecommunications and information technologies and s ervices..... such telecommunications services; (A) are essential to education, public health, or public safety; (B) have, through the operation of market choices by customers, been subscribed to by a substantial majority of residential customers; (C) are being deployed in public telecommunications networks by telecommunications carriers; and (D) are consistent with the public interest, convenience, and necessity........"
Universal service has been in effect de facto since the Kingsbury decision of 1913. 2 This implicitly allowed AT&T to retain its monopoly subject to the agreement to provide, ultimately, universal service. The universal service would mean that there would be access to all people to telephone services and that for poor people that servi ce would be subsidized. The state PUCs then followed up on this and embodied this in state regulatory requirements. In effect, AT&T and the BOCs were transferring wealth fro the "rich" to those who could not pay for such services, either bec ause of their income or because the costs to provide services to that individual would be prohibitive. This was then an enforced payment, established and managed by the BOCs, for the purpose of collecting moneys from the haves for redistribution by the BO Cs to what was perceived as the have nots. Needless to say this is per se taxation. From a Constitutional perspective such rights inure solely to the states and the Federal governments and under the Commerce Clause it is highly problematic that any indepe ndent third party has any right to tax especially as regards to interstate commerce. Needless to say there has never been a challenge her.
The Universal services fund was and still is a taxation by the BOCs to redistribute income. It also is a pool of funds to be used by them as a vehicle to bar competition. The universal services issue however goes to the heart of the interconnection issue . The RBOCs have used this ruse as a means to control competition in two ways. First, in interexchange access they have charged an access fee disproportionately higher than costs since it was then used as a basis for universal services. This was the taxat ion issue. Second, they have used a unilateral fee for any other interconnect player. Thus cellular companies, arguable providing local services, pay for initiating ad terminating calls. This has been changed by the new Act.
The Act has mandated a separate Universal Services fund to be managed by the Government, and thus the Governments powers to tax are valid and this is a legal act in contrast to the arguably illegal actions of the RBOCs in the pursuit of taxation. Second, the Act mandates balanced interconnection.
To better understand where the legal applications will be addressed we first present an overview of the major theories behind the applications of the antitrust laws. This will be important since these theoretical basis are not only applied to antitrust l aw but also to the enactment of the administrative regulations in the application of the Telecommunications Act. The litigation of any case in this area will require an understanding of the philosophical framework underlying its application.
2.1 Philosophical Underpinnings
The issues of political philosophy may seem a far cry from wireless communications but it is clearly in the middle of it. Any process which provides a service which the government is in the middle of will perforce have a political element and in turn an o verriding political philosophy. We consider two philosophies and their implications.
The first is the Rawls philosophy of John Rawls. His philosophy has three elements. The first is his concept of an Original Position. The Original Position is that all governments are based on a "contract" between its citizens and that the idea l contract is one developed in a consensus between all its citizens that allow it and them to agreement on principles of government. This is like Rousseau and the Social Contract. It is a contract amongst and between the citizens and the government, one a nd indistinguishable. From this follows the two Rawls principles of justice; First Principle, each persons shall have equal rights to the most extensive total system of equal basic liberties with a similar system of liberty for all, and Second Principle, social and economic inequalities are to be arranged so that they both, (I) provide the greatest benefit to the least advantaged, and (ii) attached to offices and positions open to all under conditions of fair equality of opportunity. 3
One may say what does this have to do with the Internet. Simply stated this philosophy controls access prices and who "must have" access. As to access prices, this is reflected in the Baumol Willig theorem of access pricing. They have used the c oncept of Ramsey pricing, also know as second best pricing. This is a sub-optimal version of Pareto pricing. Pareto pricing is a pricing mechanism in the market whereby any change in one person to increase their welfare will not diminish the welfare of an y other person. Thus something is Pareto optimal if I give you one more candy bar, that increase your welfare or happiness, and that their result of doing so does not upset anyone else. Hardly a reasonable assumption but a key basis of economic. The Ramse y scheme tries to balance welfare and profit.
The Baumol Willig theorem states that we want to maximize the welfare of the populace while keeping the profits of the monopolies high. This is a classical example of an ad hoc propiter hoc theorem. Clearly the result is that we tax the people and subsid ize the monopoly.
The other issue is how do we measure welfare. If we are a Rawlsian then we measure welfare as the welfare of the least of us and not the average welfare. Rawls states that if we maximize average welfare then we disadvantage the least of us and this is no t just. Thus as a Rawlsian we demand Universal Service. We must insist that all people have access to all service elements, whether it makes economic senses or not, we do so via wealth transfer.
Hopefully, this political theory should now not seem too foreign. Ralwsians favor the implementation of access fees and the implementation of Universal Service. Indeed, the true Rawlsian would impute Universal Service to even computer terminals as has be en stated by Vice President Gore.
In contrast is the classic liberal, now called libertarian view. It is more a combination of minimal government involvement and maximizing utility to the consumer. This is the philosophy of the utilitarian. Here we assume that government has a de minimis role and that the market follows of its own accord and that the market, in an Adam Smith fashion, will clear any inefficiencies of distribution and pricing mechanisms. It assumes that each business should stand on its own stead and that utility is maximi zed on average. The result from the libertarian school, as opposed to the contractarians or Rawlsians, is the elimination of access fees and the elimination of universal Service.
It will be important to recognize that these political philosophies dominate the overall play of regulation is all markets. These two schools of thoughts, the libertarians versus the contractarians, whether they know they are one or not, will have a grea t deal to do with our development as an industry.
2.2 Rawlsian Approach
Rawls has proposed a theory of justice that is a statement of what many proponents of antitrust theory ion the mid-fifties and sixties promulgated. The essence of Rawls' theory has three elements; 4
Original Proposition: There exists a means and method for a society to establish a Contract amongst and between themselves. This Contract thus created in this society of the just is one that maximizes the return on every transaction to the least of the individuals in the soci ety. 5 This approach to contractarianism is one related to individuals in a non-bargaining environment establishing between and amongst themselves a "contract" to govern their society. 6 There are two elements contained herein. The first is the essence of a contract, and in fact a form of social contract between the members of society and amongst them as a whole. The second element is that of a view towards man as a constrained and uncons trained view of human nature. 7 The unconstrained view states that man, individually and in concert, has the capabilities of feeling other people's needs as more important than his own, and therefore we all act impartially, even when the individuals own interest are at stake. The constr ained view is to make the best of the possibilities which exist within the constraint.
For example, the constrained view of universal service is one which would state that if it costs a certain amount to provide the service, an there is a portion of the society not able to purchase the service, then there is no overriding need to provide i t if such a provision is uneconomical and places a significant burden on the other member of society. The unconstrained view, as a form of socialism, states that if there is the least of us in want for whatever the telecommunications revolution has in sto re, then they should have access to it at whatever cost. 8 One can see that the current trend in Universal Service is such an unconstrained view, especially as viewed by the current Vice President in his actions over the past four years.
Rawls approach to this contract is one wherein the individuals in the society collect themselves as individuals, and agree to a plan for the operations of that society.
First Principle of Justice: each person shall have equal rights and access to the greatest set of equal fundamental personal liberties.
Second Principle of Justice: social and economic inequalities are to be arranged so that they both, (i) provide the greatest benefit to the least advantaged., and (ii) attached to positions available to each individual under conditions of fair equality of opportunity.
This latter elements is the means to establish a Schupeterian form of socialistic control. If we were to define the public welfare by a function W, and each individual listed as a variable I n , then the policy choice, P k , is chosen such that the welfare is maximized for the least advantaged. Specifically, if the utility of policy P k to I n is a function U, then we defined a Rawlsian system as one which performs the following mathematical function: 9
max W(u 1 ,.....u N : P 0 ,P M ) + l S [U(I n : P M ) - U(I n : P 0 ) ]
where P 0 is the initial state and P M is the application of the new policy. What this states is that we want to maximize the society welfare subject to the constraint that no individual suffers due to the change.
We can compare this to the utilitarian school which states that we seek the maximization of:
max average W(u 1 ,.....u N : P 0 ,P M )
which is not constrained on what happens to any one individual but to society as a whole. This approach is also one applied by Baumol and Willig in the establishment of interconnections and access rates and is the basis of the Ramsey tax policy. The Rams ey approach is Rawlsian whereas the approach of an Adam Smith or other utilitarians is the average approach. 10
The application of the Rawlsian theory of justice is a key factor in the current telecommunications act. Specifically it is an element of the universal service portion in that the Act requires that the least of us receive the same as the greatest of us. 11
2.3 Habermas Theory
A slight distinction to Rawls is the theory of Habermas. In the Rawlsian case the contract is generated in a mass meeting of all people. Rawls assume that such a meeting would engender the development of a justice system that would provide for the equal j ustice for all and ensure the result that the least of all would be protected. Habermas, and as recently stated in his most current publications, takes a similar but drastically different tract. He assume some form of representative government which comes up with a more average form of justice. One could state that this is what we see in many republican democracies. In contrast to least getting the same, in the Habermas system one would expect a more averaging of the welfare function over the population.
The Habermas distinction is that the contract is created by a different and representative group which takes into account certain economic realities. In some sense this is a representative democracy. The Habermas approach is in many ways intermediate to Rawls and Bentham.
2.4 Utilitarian School
The utilitarian school has as one of its key developer Jeremy Bentham. Bentham, to paraphrase Posner, states that people are rational maximizes of their own utility or satisfactions in all areas of life and that economic efficiency is an ethical and scien tific concept. 12 The utilitarian approach does not generally focus on the individual, and thus that constraint of Rawls is absent. It assumes that the individual can make an economic choice. For example, as regards Universal Service, if I decide to live in Montana in the mountains, I have made a utilitarian choice of maximizing my satisfaction.
In that choice, independent of government intervention, I have chose to forgo the advantages of a broad based telecommunications access. I will not have a Mbps link to my cottage, I will not have video on demand, and I may not have access to the Internet . I have no social contract with others and deal solely with myself. That choice is then my choice. If however, the Government is a Rawlsian approach, mandates that I have the telecommunications access, then this may of may not be reflected in my choice o f where to live but it clearly costs the other members of society who are now taxed to pay for this added satisfaction, albeit questionable, which I am now the recipient of.
As a utilitarian policy analyst, I would not require any form of Universal Service, and in particular I would argue for free and competitive open markets. These two extremes will be at the hear of the battle over Universal Service.
3. LMDS
LMDS, Local Multipoint Distribution Services, is a generic terms for the deployment of integrated two way voice, video and data services using the generic 28 GHz band. 13 The LMDS systems are to deployed after the FCC auctions the spectrum in this band. This is expected in late 1996 or early 1997. LMDS is an immediate extension to MMDS, microwave multipoint distribution services. LMDS may very well supplant MMDS. MMDS in m any ways is merely microwave to buildings and is in reality a real estate business since the target market is multiple dwelling units. However, LMDS it has been argued is focused on all forms of residential and commercial applications.
The service provided are a mix of video, voice and data. The system must have the capability to reallocate the amount of capacity between all three general areas. This allocation must be both on a quasi-static basis as well as adjustments on a real time basis. The system thus has a dynamic management capability that it must provide.
3.1 Services
The system may provide, at a minimum, the following general services:
Voice : The system may provide full switched toll grade quality voice service. The voice quality may be telephone toll grade or better and there may be no delays in speech that are perceptible to the user. The user may interface with the system by a standard meth od or means typically being an RJ-11 standard telephone jack employing their own standard telephone in the case of a residential user. The voice user is not expected to change any of their infrastructure interfaces. The "normal" telephone connec tion may be provided by means of the LMDS local interface unit, the LIU. The LIU may be compatible with any and all normal accepted telephone interfaces. The system must also provide all typical custom calling and CLAS features as expected in normal deliv er of a competitive wire based telecommunications service.
Low Speed Data : The system may be able to provide data at the rates of 1.2 to 9.6 Kbps on a transparent basis and have this data stream integrated into the overall network fabric. The system may handle all data protocols necessary in a transparent fashion. The network ma y allow local access to value added networks from the local access point. The low speed data may be provided for over a standard voice circuit from the users premises as if there were no special requirement. There may be toll grade or better quality. The system may also be capable of support all Group 3 fax services.
Medium Speed Data : The network may be able to handle medium speed data ranging from 19.2 to 64 Kbps. The interfaces for such data may be value added network local nodes. The medium speed data may be provided for over a standard voice circuit from the users premises as if t here were no special requirement. There may be toll grade or better quality. The interconnection for 64 Kbps may also be ISDN compatible.
High Speed Data : Data rates at and in excess of 1.544 Mbps may also be provided on an as needed basis and a dedicated basis. The data rates may be between 1.544 Mbps and a maximum of 155 Mbps. The BER may be less than 10 -9 . Also it may be required to provide access to such high speed data services as Fast Ethernet and FDDI at 100 Mbps. This may require both physical layer interfaces and the datalink and network layers as specified in the particular protocol. The system mu st also support multiple layer protocols including TCP/IP. Also the data must be point to point, point to multipoint, and multi point to multi point.
Video: The network may be able to provide the user with access to analog and digitized video services. This may also enable the provisioning of interactive video services. The video services may enable a system with a minimum number of channels of 150 video cha nnels of remote programming, ten of local off-air programming, and 20 locally generated programming. The interactive video may allow for ten channels of pay per view at a minimum, and interactive channels for local information selection. Video must also s upport such tiered services as basic, premium, pay per view, and interactive. The inputs to the system are from such sources as off-air, local generated, satellite, and other sources. Sources may be analog or digital, encrypted or not.
3.2 LMDS Architecture
This section presents an overview of a possible LMDS architecture. The overall design is shown in the following Figure. Here is shown a connection to the Telcos and to external video sources. There is a connection to a Telecommunications Switching Unit, T SU, and to a Video Provisioning Unit, VPU. The connections between these and the DCU, the Digital Connection Unit, is a digital signal. The DCU takes the digital signals from the TSU and VPU and combines them in a common broadband digital signal, assigns it is a TDM form for transmission and then places it in the appropriate RF format for transmission to the BSU, the Base Station Units or nodes. The input from the BSU is also passed throughout he DCU and is fed to the TSU if it is data or voice, and to th e VPU if it is a video control signal.
The actual embodiment of these units will be left to the system integrator to complete. However, it will be essential that these be separate embodiments and separately controllable and upgradeable.
The BSU transmits in the LMDS band to the NIU in the end users premises. The signal from the BSU is TDM and the NIU return is TDMA.
The proposed carrier plan is shown in the following Figure. It shows a broadband video carrier which is TDM and a set of narrowband local video digitized inserts. It also shows a transmit voice band comprised of sets of carriers and a guard band separati ng the receive TDMA carriers. The receive carriers may also be 40 MHz or some other bandwidth occupancy. The detailed carrier plan in the following Figure shows a low band of 850 MHz and a high band of 150 MHz. The low band is further split into a video s ub-band and a voice sub band. The low band may be used for transmission from the node to the end user and the high band for transmission from the end user to the node. Allocations other than what has been proposed may be deployed depending on the ability to achieve overall system performance and services acceptability.
The BSU can use multiple frequencies by segmenting the sectors. The following Figure depicts a sectoring of 60 o sectors with hexagonal patterns. The sectors are separated by vertical and horizontal polarization as allowed expressly by the FCC. The sectors, here six, can each have higher gain antennas and can each be driven by separate systems. The added gain allow s for wider coverage. There is a balance between sectoring and increased antenna gain and the number of sectors and their cost. The larger the number of sectors the smaller the beamwidth and the greater the gain and the longer the effective range. It is a nticipated that there is an optimum sectorization depending on the terrain. In hilly areas the line of sight, LOS, is limited by obstructions so that no matter how great the gain on the antenna, and effectively how many sectors, there is a diminishing ret urn in the design. Many designs will employ a single beam or sector.
The BSU to NIU connection is from the sector antenna at the BSU to the NIU antenna which is a narrow beam and high gain antenna. The antenna beam from the NIU assures that frequency interference is kept at a minimum. This is the standard approach as was u sed in such systems as the Interdigital system. 14
4. PCS Systems
PCS, Personal Communications Services, is nothing more than the expansion of bandwidth and the introduction of competition in wireless. It is not a specific frequency band, a new technology, nor added features and services. It is the introduction of compe tition in the new markets. This is a definition that is economically driven and not technologically. One should avoid the battles between TDMA and CDMA, between 800 MHz and 1.9 GHz, between voice, and two way paging. PCS I the commoditization of air time. It will allow entrepreneurs to take unbundled air time and create highly competitive new services and systems.
When viewed in this sense, PCS is comprised of two 800 MHz providers of 30 MHz each, three 1.9 GHz providers of 30 MHz apiece, three 1.9 GHz providers at 10 MHz apiece, and a SMR set of providers in 900 MHz at about 5-10 MHz apiece. Thus there are nine w ireless providers of a commodicizable product, namely airtime. 15
The current wireless technology as embodied in the cellular communications systems is composed of several key technological elements. Specifically they are the Cell Sites, the MTSO (Mobile Telephone Switching Office) or Mobile Switching Center ("MSC& quot;), and whatever connections or management systems are in place. The connections between the cell sites and the MTSOs are digital circuits carrying the voice signals.
Unlike LMDS, PCS is mobile. PCS has less bandwidth and is generally focused on large regions of coverage. LMDS is fixed as a service, it may be delimited as a LEC to a single state, whereas PCS, as is all of cellular, able to cross state boundaries. This is a dramatically different regulatory constraint. PCS providers are Commercial Mobile Radio Service, CMRS, providers. LMDS providers are LECs.
4.1 System Elements
The design strategy in this section proposed uses a CDMA approach which has been selected since it provides the lowest cost per subscriber. The design of the system using CDMA demonstrated the ability of that technology to balance coverage and capacity. H owever, it can be shown that with the cost reductions in GSM TDMA technology the differences are getting smaller over time. We can characterize these two domains as follows:
Capacity Domain: In this case there are enough cell sites and they are deployed so that at no time is a unit too far from a cell. For high powered units this may be a great distance. For lower power units this may be quite a small distance. However, there are so many use rs per cell area that the load exceeds the capacity of a cell. To meet the demand, cells must be split and the frequency reused. This installation of new cells for the reason of reaching a capacity limit is called the Capacity Domain.
Coverage Domain : As with the capacity domain, the coverage domain is that situation when new cells are added because users are too far away from any cell. The coverage domain of PCS is that time where there are less than the saturation number of subscribers per cell in a ll of the cells. It generally is that period where the system has just been deployed and the customer base is growing.
The current analog systems were in the Capacity domain several years ago and they were predicting dire results. With the sale of portables this has shifted to the coverage domain. TDMA was a response to the capacity driver only. CDMA is a response to the coverage driver as well as the capacity driver. To balance the fluctuation between the two domains in a cost effective fashion, it is necessary to have a technological infrastructure that meets the two needs, this is provided by CDMA.
In the analog world, cell capital costs are typically $1 million per site and a typical site may cover a radius of 10 miles for a 3 watt mobile or 2 miles for a 0.6 watt portable, and a cell has a capacity of 40-50 instantaneous calls at any one time. Th e adjacent cells must use separate frequencies and thus there is a proliferation of cells and a significant amount of capital in cell sites. For example, New York has over $350 million in cell sites per carrier and Boston has over $150 million. This numbe r is for each carrier, A and B side. Thus in New York, at $250 million and for 125,000 subscribers, this is $2,000 of cell capital per subscriber.
Capital in the system will be divided into two major categories; local service infrastructure (LSI) and national service infrastructure (NSI). The LSI portion consists of all elements of the system up to and including the switch. The NSI portion is all e lements from the switch on back. The NSI will also include elements that comprise the databases and computer support.
Six independent elements have been identified in the overall PCS architecture. They are as follows;
CDMA dramatically changes the cost equation. First, CDMA will allow 10 to 20 times the capacity per unit bandwidth as analog. Second, CDMA allows for use of the same frequencies by keeping separations through the direct sequence codes in the technology. S imply put, a CDMA cell site may cover a 3 mile radius, yet have the capacity of 75 channels per 1.25 MHz of Bandwidth. Using a CDMA system, one may cover a greater area and thus be run at maximum utilization of close to 90% or more per site.
The costs associated with this configuration are those capital and operating costs of the cell and the MTSO as well as the carriers charges for the PSTN. For a cell there are specific life cycle factors that control its overall costs are: Cell Site Locat ion and Planning, Cell Construction, Cell Capital, Cell Installation, Cell Operations, Cell Maintenance, and Cell Repairs.
The first four items are part of the initial capitalization and may take anywhere from six months to two years, depending on how quickly access is allowed to the site. The last three elements are ongoing. In some systems, the sum of all these costs for t he full life of a cell, seven to ten years, may be two to three times the cell installation capital. Thus cell site life cycle costs are a critical factor to manage in a system.
The following Figure depicts the PCS architecture used for pricing.
4.2.1.1 Architecture of PCS
To effectively compare technological alternatives we must have models for the effective utilization of capital in the two cases. In this section we shall develop these models in summary form. We assume that the system is composed of the following three ge neric elements;
Base Terminal Stations (BTS): These devices are placed in the field and there are as many BTSs as are need for either coverage or capacity. The first demand is coverage. A BTS may cover 20-30 square miles, depending on the power, the modulation, the multiple access, and the capabilit ies of the wireless end user terminal. For example, in CDMA with PCS, a BTS has three sectors, and covers three mile radius or about 33 sq. mi. per BTS. If there are no customers, then for 1,000 sq. mi., one need approximately 30 BTSs. A BTS also serves o ne or more CDMA channels. In narrow Block CDMA (namely 1.25 MHz per channel), the CDMA channels must be added each time the system load goes beyond the capacity of one link. Namely, a CDMA channel at 1.25 MHz and with three sectors services 75 instantaneo us channels or "trunks", whereas analog services 7. 16 If a user is busy only 100 minutes per month, then this is an activity ration of 1%, thus 75 trunks handle 7,500 subscribers in this 30 sq mile area. If there is a 10% penetration, then the population is 75,000 people, or PoPs, in 30 square miles, or abou t 2,700 PoPs per square mile. This is a high population density. As the traffic increase, more CDMA channels must be added. Also in any system, trunk interfaces are added as the trunks are added, perforce of traffic growth.
Base Station Controllers (BSC): The BSC provides for the overall coordination and processing of the switched signals. It typically can handle a multiple set of BTSs and a multiple set of trunks. In the current CDMA narrowband system, a BSC handles up to 50 BTSs.
Switches (SW): The switch for Mobile Switching Center ("MSC"), interfaces with the LECs and the IECs. It is sized based on a fixed component and a component dependent upon the number of trunks. Newer systems use ATM switching which has proven to be more effic ient for the packet type voice signals integrated with data in a wireless environment.
The financial models for a narrowband CDMA system is presented below. It assumes that there are 1.25 MHz channels along with a total available spectrum as discussed above, and it assumes that the area covered is 1,000 sq. mi. The results show Capital per subscriber as a function of the total subscriber base. It should be noted that there is significant scale in the lower end.
The following set of sizing are based upon vendor supplied financial numbers but are retail and do not include any volume discounts or other factors. Note that the system capital for the 10 MHz system is about $366 per sub and reaches that at almost 50,0 00 subs as we have specified. From that point on Capital per sub is all marginal, namely it lacks scale. 17
Note in the second case, whether we have 30 MHz, we have reduced Capital per subscriber from $366 to $336. This is a $30 per subscriber penalty for only 10 MHz but may be more than set aside by the lower cost of the spectrum.
4.2.1.2 CDMA (1.25 MHz Channels, 10 MHz Spectrum)
4.2.1.3 CDMA (1.25 MHz Channels, 30 MHz Spectrum)
CDMA has a larger single cell radius at 0.6 W than does all of the other systems. This is due to the lower E b /N o needed for the link. This will have a dramatic effect in achieving the targeted cost per customer number. We shall use the example of CDMA technology to demonstrate how this new technological infrastructure can enable the new market. We shall briefly des cribe the CDMA system and then proceed to the financial implications of using this new technology.
4.3 Operations Support Systems
The OSS elements are generally computers, workstations, memory units and other MIS type systems. Capital is composed of initial fixed capital and then incremental growth capital. These have been sized and are part of the overall model. We will show their impact when we develop the design of the system.
The provision of OSS will entail several dimensions of service capabilities. These may or may not be from a single service providers but must be able to be integrated into a single service provisioning element. The following are the sets of functions to be provided :
Network Management
The local and national backbone network must be managed and controlled in a real time fashion. Operating entities, at all levels of operation, must have the capability of being monitored as to operational effectiveness, network performance, and impact on their interconnecting network elements. The Network Manager must be able to determine the locations of any and all outages or system degradation points in the network, or in any other network that a customer may have access to.
IEC Interface Management
IEC Management must be performed to ensure the establishment and proper maintenance of any and all IEC interfaces and connections to the local PCS network. The overall management service will include such items as circuit ordering and scheduling, circuit interface negotiations, optimization of network design, and the physical management of the integration of the networks. It has been assumed that the IEC interfaces will be consistent with all other equal access provisions and that no IEC will receive any preferential treatment.
Customer Service
The Customer Service function will provide customer service capabilities supporting such areas as billing, service quality, inquiries, service features, service upgrades, and complaints. Customer Service is the most important part of the provision of serv ice. The customer only needs Customer Service when the service is not totally transparent and thus when the service is not meeting the customers needs. Therefore, Customer Service is the MOST critical function that can be provided and must be provided wit h utmost care and effectiveness.
Billing
The Billing Function must be responsible for the full life cycle factors associate with billing. This includes the capture of billing data, both local and IEC, the processing of the data, the preparation of the bill, the issuance of the bill, and the coll ection, reporting of and corrections to the bill. The billing function in essence consists of all functions necessary to collect the bill for services rendered, commencing from the time the service is requested, through the necessary intermediates steps a nd through all intermediaries.
Roaming Implementation
The Roaming functions are required to provide a national and seamless service. The roaming functions require the establishment of a national database and a national identifier system. All portables must have an identifier and self registration facility to identify themselves as they enter a new system. This must then be integrated into the active roaming database and all calls must be routed accordingly.
Repair Dispatching and Maintenance
The RD&M function is required when a fault is detected. The function prepares the trouble ticket and the dispatch ticket and the inventory dispatch ticket. It closes out all repairs and reports on the results.
Inventory Management (MRO/MRP)
The Inventory Management function, also providing Materials Resource Planning (MRP) and Material Resource Ordering (MRO) functions, will be responsible for the ordering and inventorying of all system and network elements needed for growth, spares, and mai ntenance. The function must be fully integrated and electronically supported ensuring the minimum response time and cost for inventory carrying. As a goal, the Manager seeks to have a "Just in Time" system that ensures the availability of the pa rts needed without the need for any stockpiling of equipment. This not only applies to the network elements but to the portables sent to the customers as well.
Operator Services and Directory Assistance
The Operator Services and Directory System intended to support access to all PCS customers. This system must allow any individual in any location to obtain ready access to any PCS subscriber. The objective is to ensure that all calls are equally inbound a nd outbound.
4.4 Operations Expenses
The operations of a PCS system, or any telecommunications system, for that matter, has intrinsically several costs to be included. We divide these costs into the following categories;
Cost of Goods : The costs associated with the provision of materials that may be held in inventory. We shall consider in this case that provision of the terminal, namely the portable, as a cost of goods. We have assumed a terminal cost of $200.
Cost of Service : This will be the costs associated with the access fees. We have assumed zero access fees throughout. It should also be noted that we have assumed that we are not charging AT&T or the other IECs with an access fee. This will be $0.05 per minute. This will be used to compete against the LEC.
Cost of Sales: This is the cost of all of the elements of acquiring and maintaining the customer. We have assumed that all costs are about $200 per new customer or 15% of the gross revenue.
Cost of Operations : This includes the LSI Operations as well as the OSS and are estimated at about $8 per month per subscriber.
The following three Table depicts the capital models for this type of PCS operations.
4.4.1.1 PCS Capital Plant
| Year |
|
1996 |
1997 |
1998 |
1999 |
2000 |
2001 |
2002 |
2003 |
2004 |
2005 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Switch Capital |
|
$40,000 |
$100,000 |
$100,000 |
$100,000 |
$130,000 |
$245,000 |
$315,000 |
$375,000 |
$430,000 |
$515,000 |
|
BTS Capital |
|
$27,803 |
$112,811 |
$175,981 |
$245,896 |
$456,200 |
$685,097 |
$817,893 |
$947,176 |
$1,067,454 |
$1,311,769 |
|
BSC Capital |
|
$6,273 |
$36,906 |
$48,514 |
$70,985 |
$152,819 |
$222,148 |
$263,401 |
$303,563 |
$340,928 |
$416,825 |
|
CPU Capital |
|
$1,000 |
$1,000 |
$1,000 |
$1,000 |
$1,000 |
$1,000 |
$1,000 |
$1,000 |
$1,000 |
$1,000 |
|
Cell Interconnect |
|
$5,370 |
$20,400 |
$28,620 |
$32,700 |
$36,810 |
$40,890 |
$40,890 |
$40,890 |
$40,890 |
$40,890 |
|
Switch Interconnect |
|
$1,600 |
$4,000 |
$4,000 |
$4,000 |
$5,200 |
$9,800 |
$12,600 |
$15,000 |
$17,200 |
$20,600 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Total New Capital |
|
$82,046 |
$193,071 |
$82,997 |
$96,466 |
$327,448 |
$421,907 |
$246,849 |
$231,846 |
$214,843 |
$408,612 |
|
Accum Capital |
|
$82,046 |
$275,117 |
$358,115 |
$454,581 |
$782,028 |
$1,203,935 |
$1,450,784 |
$1,682,630 |
$1,897,473 |
$2,306,084 |
|
Deprecation |
|
$8,205 |
$27,512 |
$35,811 |
$45,458 |
$78,203 |
$120,393 |
$145,078 |
$168,263 |
$189,747 |
$230,608 |
|
Accum Cap-Accum Dep |
|
$73,842 |
$239,401 |
$286,587 |
$337,595 |
$586,839 |
$888,352 |
$990,123 |
$1,053,706 |
$1,078,802 |
$1,256,805 |
|
Accum Dep |
|
$73,842 |
$313,243 |
$599,829 |
$937,424 |
$1,524,263 |
$2,412,616 |
$3,402,739 |
$4,456,445 |
$5,535,247 |
$6,792,052 |
|
Capital per Sub (Eff) |
|
$2,532.85 |
$1,263.07 |
$668.54 |
$491.76 |
$427.84 |
$400.86 |
$373.38 |
$362.05 |
$353.14 |
$358.47 |
|
Depreciation Per Sub (Eff) |
|
$253.29 |
$126.31 |
$66.85 |
$49.18 |
$42.78 |
$40.09 |
$37.34 |
$36.20 |
$35.31 |
$35.85 |
5. Economic Comparisons
This section presents several of the general economic factors that give the value per PoP to a PCS property and an LMDS property. The assumptions on both cases are similar:
5.1 PCS Valuations
The issues of PCS valuations is based on several factors:
The following Table depicts the value per PoP depending on total market size and terminal penetration in year ten. It should be recalled that bidding is at 50% of NPV. The following Figure depicts this value per PoP in terms of the same factors.
The following Table depicts typical values for BTAs that were bid upon their total PoPs and bid values,
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1,123,678 |
$32,405,632 |
$28.84 |
$14.42 |
$16,202,816 |
|
|
|
978,311 |
$27,018,810 |
$27.62 |
$13.81 |
$13,509,405 |
|
|
|
357,482 |
$8,008,615 |
$22.40 |
$11.20 |
$4,004,308 |
|
|
|
4,133,895 |
$223,745,901 |
$54.12 |
$27.06 |
$111,872,951 |
|
|
|
204,256 |
$4,313,019 |
$21.12 |
$10.56 |
$2,156,509 |
|
|
|
139,352 |
$2,866,548 |
$20.57 |
$10.29 |
$1,433,274 |
|
|
|
672,970 |
$16,859,882 |
$25.05 |
$12.53 |
$8,429,941 |
|
|
|
709,705 |
$17,999,199 |
$25.36 |
$12.68 |
$8,999,599 |
|
|
|
316,838 |
$6,989,902 |
$22.06 |
$11.03 |
$3,494,951 |
|
|
|
221,697 |
$4,713,778 |
$21.26 |
$10.63 |
$2,356,889 |
|
|
|
471,614 |
$11,017,638 |
$23.36 |
$11.68 |
$5,508,819 |
|
|
|
86,936 |
$1,750,044 |
$20.13 |
$10.07 |
$875,022 |
|
|
|
165,671 |
$3,444,571 |
$20.79 |
$10.40 |
$1,722,286 |
|
|
|
111,709 |
$2,271,977 |
$20.34 |
$10.17 |
$1,135,989 |
|
|
|
167,576 |
$3,486,861 |
$20.81 |
$10.40 |
$1,743,430 |
|
|
|
540,704 |
$12,945,488 |
$23.94 |
$11.97 |
$6,472,744 |
|
|
|
1,509,789 |
$48,437,394 |
$32.08 |
$16.04 |
$24,218,697 |
|
|
|
369,128 |
$8,305,629 |
$22.50 |
$11.25 |
$4,152,815 |
|
|
|
97,987 |
$1,981,600 |
$20.22 |
$10.11 |
$990,800 |
|
|
|
12,379,298 |
$438,562,490 |
|
|
$219,281,245 |
5.2 LMDS Economics
The economics of LMDS follow a similar path as does that of PCS. The differences are several:
The following Figure presents the NPV/PoP as a function of the average capital per subscriber.
Those numbers are generally comparable to the PCS numbers. The bidding for these properties, again being based on BTAs are expected to be late in 1996 of sometime in 1997.
6. Interconnection
The interconnection issue is a major factor in the deployment of any wireless systems. The new Act provides a significantly changed platform upon which the new entrants may operate. This section provides an analysis of the interconnect problem from the CM RS's facilities to the I-LEC, the incumbent Local Exchange Carrier and from a C-LEC to the I-LEC. As has been discussed in previous papers, the interconnect issue for a wireless carrier falls into two categories; intra-plant and inter-plant. 18 The intra-plant issue is that between cell sites and the carriers own switch and the inter is between the carrier's switching facilities and the I-LEC's facilities. The intra was discussed in Telmarc Report TR-96-008. The overview of these issues is shown below.
This Figure depicts three issues: first is that the intra-plant facilities are generally under the total control of the carrier. Second, that the end office I-LEC interconnect is clearly under the control of the Section 251 reciprocal compensation rule. N amely, such agreements as those between WinStar and NYNEX allow for termination of traffic here on a mutual compensation basis. Third, the real problem is how does one get from a single MSC, to several access tandems and then ultimately to dozens of end o ffices. This report addresses those issues.
The overall goal of this report is twofold. First to address the technical issues related to the interconnection, especially what options are available to tandem interconnection. Second, what are the resultant regulatory options that may be available to the carrier.
Any new carrier must be aware of these options before they interconnect since these interconnection options present significant fixed costs to the carrier and there may be ways to move these monthly fixed costs into some variable form or to move them int o a form of carrier owned facilities.
6.1 C-LEC versus CMRS
In this section we develop a detailed review of the new regulatory structure as applied to interconnect. There is a difference between a C-LEC, a competitive LEC, and a CMRS, commercial mobile radio services provider. These are the two classes of players interconnecting under Section 251 of the Act.
A C-LEC is a non-incumbent LEC. An incumbent LEC is generally a RBOC. A LEC is defined by the Act as:
"LOCAL EXCHANGE CARRIER- The term "local exchange carrier" means any person that is engaged in the provision of telephone exchange service or exchange access. Such term does not include a person insofar as such person is engaged in the prov ision of a commercial mobile service under section 332(c), except to the extent that the Commission finds that such service should be included in the definition of such term."
The definition of telephone exchange services and exchange access services is as follows:
"EXCHANGE ACCESS- As per the Act, Sec.3(b)(2), the term Exchange Access means the offering of access to telephone exchange services or facilities for the purpose of the origination or termination of telephone toll services."
and,
"Telephone Exchange Service - Telephone Exchange Service is defined in 47 U.S.C. Sec. 153 (r)means service within a telephone exchange, or within a connected system of telephone exchanges within the same exchange area operated to furnish to Subscribe rs intercommunicating service of the character ordinarily furnished by a single exchange, and which is covered by the exchange service charge."
Exchange Services is generally the provision of toll telephone services whereas telephone exchange services is local services directly to the end user or customer.
In contrast a CMRS is defined as:
"(i) CMRS: A Commercial Mobile Radio Service ("CMRS") as defined by 47 U.S.C. Section 332 and from the Code, Section 153 (n). Specifically, Commercial Mobile Radio Service means any mobile service (as defined in section 47 U.S.C Section 153(n)) that is pr ovided for profit and makes interconnected service available (A) to the public or (B) to such classes of eligible Users as to be effectively available to a substantial portion of the public, as specified by regulation by the Federal Communications Commiss ion."
"(ii) MOBILE SERVICE : As defined in section 47 U.S.C Section 153(n), Mobile Service means a radio communication service carried on between mobile stations or receivers and land stations, and by mobile stations communicating among themselves, and includes (1) both one-way and two-way radio communication services, (2) a mobile service which provides a regularly interacting group of base, mobile, portable , and associated control and relay stations (whether licensed on an individual, cooperative, or multiple basis) for private one-way or two-way land mobile radio communications by eligible Users over designated areas of operation, and (3) any service for which a license is required in a personal communications service established pursuant to the proceeding entitled ''Amendment to the Commission's Rules to Establish New Personal Communications Services'' (GEN Docket No. 90-314; ET Docket No. 92-100), or any successor proceeding.&quo t;
The mobile service definition requires three elements; two way communications, over a an infrastructure and that the operator is in possession of an FCC license to provide such services. There is an exception as stated by the FCC for a CMRS, namely as rel ates to a reseller of CMRS services. Specifically, the FCC has ruled: 19
"Finally, we conclude that mobile resale service is included within the general category of mobile services, as defined in Section 3(n) and for purposes of regulation under Section 332, since resale of mobile service can only exist if there is an und erlying licensed service. There is no indication in the statute or the legislative history that resellers are not "mobile service" providers or exempt from the Section 332 regulatory classification, and we see no reason to establish such an exem ption."
This simply states that even reseller are CMRS and thus also must be concerned with this issue.
6.2 Rates for Interconnection
The rates for interconnection have been established in the FCC First R&O. However this has been set aside by the Eight Federal District Court until it is reviewed. However, many f the RBOCs have already entered into interconnect agreements or are curr ently negotiating them. This section presents a comparison between several players in the market and presents the current pricing schedules.
The following Table compares the LEC status to that of a CMRS. This report focuses on the CMRS advantages
|
|
|
|
|
|
|
|
|
|
|
Coverage |
IntraState | IntraState | MTA 20 |
| Jurisdiction | State PUC | State PUC | FCC |
| Interconnection | Section 251 | Section 251 | Section 251 |
| Reciprocal Compensation |
Mandated
CFR §51.717 |
|
Allowable
CFR §51.717 |
| Bill and Keep | Optional | Optional | Optional |
|
|
|
|
|
| Resale | Mandated | Mandated | NA |
| Reciprocal Compensation | Mandated | Mandated | NA |
| Dialing Parity | Mandated | Mandated | NA |
| Access to Rights of Way | Mandated | Mandated | NA |
|
|
|
|
|
| Duty to Negotiate | Required | NA | NA |
| Unbundling | Required | NA | NA |
| Co-Location | Required | NA | NA |
| Interconnection | Required | NA | NA |
The requirement by the new CFR is related to local termination traffic. This is defined as:
" Local Telecommunications Traffic means: (1) telecommunications traffic between a LEC and a telecommunications carrier other than a CMRS provider that originates and terminates within a local service area established by the state commission; or (2) telecommunications traffic between a LEC and a CMRS provider that, at the beginning of the call, originates and terminates within the same Major Trading Area"
The latter statement is of significant import to a CMRS carrier. It covers all of an MTA and since the New York MTA covers eastern New York, New Jersey, Vermont, and eastern Pennsylvania, it is a significant advantage over any LEC. The default tariffs app lied by the FCC in the new CFR are as follows:
|
|
|
|
|
|
|
|
|
Symmetrical Reciprocal |
Yes | CFR §51.711 |
| Termination of Local Traffic | No less than $0.002 per minute and no more than $0.004 per minute | CFR 51.707(b)(1) |
| Transport of local traffic | Same as in Termination and in Tandem |
CFR §51.707(b)(1),
and §51.513(d)(3),(4),(5) |
| Tandem Switching | $0.0015 per minute of use. | CFR §51.715(b)(3) |
The actual interconnect agreement negotiated between NYNEX and WinStar reflects the following rates. It should be remembered that although WinStar is a wireless carrier it is not a CMRS, it is a LEC. It is a C-LEC and thus there are certain distinctions. Also, all three are common carriers, namely the I-LEC, the C-LEC and a CMRS.
|
|
|
|
|
|
|
|
|
Termination of Local Traffic |
|
$0.0080 per minute |
| Transit Service |
|
$0.0035 per min |
| Number portability |
$1 per month residential
$2 per month business $20 per ported number |
|
| Unbundled ports | $8 per month |
|
| 911 Connections | $252 per month per DS1 plus $100 per month per voice grade trunk activation. |
|
| Directory Assistance |
$0.32 per message; branded
$0.57 with DAC |
|
|
Reciprocal Rule Rate per minute
Peak Rate = $0.009 per min OffPeak Rate = $0.0065 per min OPM=originated peak min OOPM orig. OffPeak min |
[Total Carrier Min+Total ILEC Min} |
|
The remainder of this paper presents the detail regarding the regulatory, technical and operating issues of inter carrier interconnection.
7. Conclusions
This paper has presented a detailed analysis of the costs of two wireless schemes; PCS and LMDS. The similarities are greatest as are the difference. PCS is a bandwidth limited system, 30 MHz of bandwidth, at 1.9 GHz center frequency. LMDS is 1 GHz of ban dwidth at 28 GHz center frequency. One is longer range, although not really that great, and the other is greater bandwidth, but the issue is for what purpose. The underlying question is; is there a sustainable business in wireless or has a glut been creat ed? Also the second question is, if universal service is a socially acceptable and required goal, then which of these technologies is the most resource effective in deployment.
We argue that any single service, if there is a market, can be competitive. We further argue that wireless may not be the most efficient service for large areas, especially when the population density falls below 200 PoPs/Sq mi. This is especially true i n rural areas. Possibly point to point may work but even this is debatable.
The major concern is the general assumption that bandwidth is a scarce resource. This is true if can not been efficiently used. In the current models we see that with PCS alone we can service the entire US telephone network several times over. This may a ctually mean that the total capital deployed in PCS may be a poor use of capital resources on a national scale. On the other hand this excess capacity in a free market will drive prices down and drive for newer and more creative applications. The missing element has always been the load that data may apply on the system. This may be the savior for PCS.
As for Universal Service, the FCC seems to be focusing on the use of a fund type approach, taking the "taxation" control from the RBOCs. This may be the most effective mechanism. The issue that may academic policy analysts have tried to raise i s that a wireless provider, in return for the right given them by the government, have a corresponding right to "play fair" with the monopolists and be required to provide universal wireless service. This bizarre set of reasoning goes beyond the wildest dreams of Schumpeter. First, the PCS entrants have risked billions to buy PCS spectrum, which is more than can be said for the monopolist RBOCs, who in may cases usurped their properties at the turn of the nineteenth century. Secondly, if economi c efficiencies are of any importance, why not let the market determine who is the most efficient provider. Thirdly, why must the monopolist be protected. In fact if telecommunications is truly commodicized with wireless, then as we have seen in long dista nce, the market is very efficient without any governmental strictures such as compulsory universal service.
8. References
1. Areeda, P., L. Kaplow, Antitrust Analysis, Little Brown and Co (Boston), 1988.
2. Baumol, W.J., Economic Theory and Operations Analysis, Prentice Hall (Englewood Cliffs, NJ), 1972.
3. Baumol and Sidak, Toward Competition in Local Telephony, MIT Press (Cambridge, MA), 1994.
4. Bork, R.H., The Antitrust Paradox, Free Press (New York), 1978.
5. Brown, S.J., D.S Sibley, The Theory of Utility Pricing, Cambridge Press (Cambridge, England), 1986.
6. Coll, S. The Deal of the Century, Atheneum (New York), 1986.
7. Dertouzos, M.L., J. Moses, The Computer Age, MIT Press (Cambridge, MA), 1979.
8. de Sola Pool, I., Technologies Without Barriers, Harvard University Press (Cambridge, MA), 1990.
9. de Sola Pool, I., The Social Impact of the Telephone, MIT Press (Cambridge, MA), 1977.
10. Dworkin, R., Law's Empire, Harvard University Press (Cambridge, MA), 1986.
11. Kronstein, H. et al, Modern American Antitrust Law, Oceana Publications (New York), 1958.Fisher, F.M., Antitrust and Regulation, MIT Press (Cambridge, MA), 1985.
12. Habermas, J., Between Facts and Norms, MIT Press (Cambridge, MA), 1996.
13. Hayek, F.A., The Road to Serfdom, University of Chicago Press (Chicago, IL), 1994.
14. Hoffman, M.A., A.I. Winard, Antitrust Law and Techniques, Matthew Bender (Albany, NY), 1963.
15. Holmes, O.W., The Common Law, Back Bay Books (Boston), 1963.
16. Hovenkamp, H., Antitrust, West Publishing (St. Paul, MN), 1986.
17. Hovenkamp, H., Economics and Federal Antitrust Law, West (St. Paul, MN), 1985.
18. Huber, P.W., The Geodesic Network, U.S. Department of Justice, Washington, DC, January, 1987.
19. Kahn, A.E., The Economics of Regulation, MIT Press (Cambridge, MA), 1989.
20. Kukathas, C., P. Pettit, Rawls, Stanford University Press (Samford, CA), 1990.
21. McGarty, T.P., G.J. Clancey, Cable Based Metro Area Networks, IEEE Jour on Sel Areas in Comm, Vol 1, No 5, pp 816-831, Nov 1983.
22. McGarty, T.P., Alternative Networking Architectures; Pricing, Policy, and Competition, Information Infrastructures for the 1990s, John F. Kennedy School of Government, Harvard University, November, 1990.
23. McGarty, T.P., Information Architectures and Infrastructures; Value Creation and Transfer, Nineteenth Annual Telecommunications Research Conference, Plenary Address and Paper, Solomon's Island, September, 1991.
24. McGarty, T.P., Alternative Networking Architectures, B. Kahin Editor, McGraw-Hill (New York), October, 1991.
25. McGarty, T.P., Communications Networks; A Morphological and Taxonomical Approach, Private Networks and Public Policy Conference, Columbia University, New York, October, 1991.
26. McGarty, T.P., Wireless Communications Economics, Advanced Telecommunications Institute Policy Paper, Carnegie Mellon University, February, 1992.
27. McGarty, T.P., Communications Network Morphological and Taxonomical Policy Implications, Telecommunications Policy Research Conference, Solomon's Island, MD, September, 1992.
28. McGarty, T.P., Architectures et Structures de L'Information, Reseaux, No 56, pp. 119-156, December, 1992, Paris.
29. McGarty, T.P., Economic Structural Analysis of Wireless Communications Systems, Advanced Telecommunications Institute Policy Paper, Carnegie Mellon University, February, 1993.
30. McGarty, T.P., Spectrum Allocation Alternatives; Industrial; Policy versus Fiscal Policy, MIT Universal Personal Communications Symposium, March, 1993.
31. McGarty, T.P., Wireless Access to the Local Loop, MIT Universal Personal Communications Symposium, March, 1993.
32. McGarty, T.P., Access to the Local Loop; Options, Evolution and Policy Implications, Kennedy School of Government, Harvard University, Infrastructures in Massachusetts, March, 1993.
33. McGarty, T.P., Internet Architectural and Policy Implications, Kennedy School of Government, Harvard University, Public Access to the Internet, May 26, 1993.
34. McGarty, T.P., Access Policy and the Changing Telecommunications Infrastructures, Telecommunications Policy Research Conference, Solomon's Island, MD, September, 1993.
35. McGarty, T.P., Wireless Architectural Alternatives: Current Economic Valuations versus Broadband Options, The Gilder Conjectures; Solomon's Island, MD, September, 1994
36. McGarty, T.P., From High End User to New User: A New Internet Paradigm, McGraw Hill (New York), 1995.
37. McGarty, T.P. , "Disaggregation of Telecommunications", Presented at Columbia University CITI Conference on The Impact of Cybercommunications on Telecommunications, March 8, 1996.
38. McGarty, T.P., Competition in the Local Exchange Market: An Economic and Antitrust Perspective, Federal Communications Law Journal, submitted and to be published, 1997.
39. Puucknett. T.F.T., A Concise History of the Common Law, Little Brown (Boston, MA), 1956.
40. Porter, M., Competitive Advantage, Free Press (New York), 1985.
41. Porter, M., Competitive Strategy, Free Press (New York), 1980.
42. Posner, R.A., Economic Analysis of :Law, Little Brown and Co. (Boston, MA), 1992.
43. Posner, R.A, The Problems of Jurisprudence, Harvard University Press (Cambridge, MA), 1990.
44. Posner, R.A. Overcoming Law, Harvard University Press (Cambridge, MA), 1995.
45. Posner, R.A., Antitrust Law, University of Chicago Press (Chicago, IL), 1976.
46. Posner, R.A., The Economics of Justice, Harvard University Press (Cambridge, MA), 1983.
47. Pound, R., The Spirit of the Commpn Law, Marshall Jones Co (Boston, MA), 1921.
48. Rawls, J., A Theory of Justice, Harvard University Press (Cambridge, MA), 1971.
49. Rawls, J., Political Liberalism, Columbia University Press (New York), 1996.
50. Rockefeller, E.S., Antitrust Questions and Answers, Bureau of National Affairs,(Washington), 1976.
51. Schumpeter, J.A., Capitalism, Sociualism, and Democracy, Harper & Bros (New York), 1942.
52. Shepherd, W.G., The Treatment of Market Power, Columnbia University Press (New York) 1975.
53. Sowell, T., A Conflict of Visions, Wm. Morrow (New York), 1987.
54. Spulber, D.F., Regulation and Markets, MIT Press (Cambridge, MA),1990.
55. Stocking, G.W., M.W. Watkins, Monopoly and Free Enterprise, Twentieth Century Fund (New York), 1951.
56. Sullivan, L.A., Antitrust, West Pub. (St Paul, MN), 1977.
57. Weinhaus, C.L., A.G. Ottenger, Behind the Telephone Debates, Ablex Pub. (Norwood, NJ), 1988.
| 2 |
See Weinhaus, p. 9.
|
| 3 |
See Kukathas, Rawls, Stanford University Press.
|
| 4 |
See Kukathas and Pettit for an excellent expository of the Rawlsian theory.
|
| 5 |
Indeed in the Rawlsian world the individual posits their position and does so without any negotiation and thus posits a position assuming that that individual will be the least amongst players in that society. Such a position, to create justice in that so
ciety, is a
maxi-min
position.
|
| 6 |
This is the same in many ways of the Social Contract envisioned by Rousseau.
|
| 7 |
See Sowell, pp. 18-24.
|
| 8 |
See Schumpeter, pp. 167-186. Specifically, he defines socialism as; "
an institutional pattern in which the control over the means of production and over production itself is vested in a central authority- or, as we may say, in which, as a matter of principle, the economic affairs of society belong to the public and not the
private sector
." Indeed in the case of Universal Service, the FCC and other elements of the Executive Branch have taken the property and means of production from the carriers and mandate how they are to be deployed, irrespective of an economic justification.
|
| 9 |
This is a modification of the Ramsey pricing scheme of constrained maximization. Also there are variations as discussed by Baumol are: the classic Pareto, the Kaldor, the Scitovsky, and the Bergson. The problem with these approaches is that they are all a
d hoc propiter hoc approaches.
|
| 10 |
See Brown and Sibley, p. 39. This is the Baumol Willig theorem which uses the first Rawlsian approach to maximizing the return subject to a single constraint; namely that the monopolist suffers no harm. This has been the basis of non-mutual interconnectio
n fees and has been corrected by the new Act but is currently stayed by the Sixth District Court after an avalanche of suits by the RBOCs. This is of course a different twice to Schumpeterian socialism, wherein a monopolist like the RBOC usurps the power
of the state for its own benefit.
|
| 11 |
See SEC. 254. UNIVERSAL SERVICE.
|
| 12 |
See Posner, Economics of Justice, p. 13. In addition, Posner in his other referenced works develops a [powerful theory of justice and the law and his economic analyses are powerful tools in the overall economic and policy analysis of this area.
|
| 13 |
This is from the FCC R&O on LMDS, FCC Fourth Report and Order, CC- Docket No. 92-297, July 17, 1996. This may be amended from time to time and the compliance with the FCC rules and regulations is incumbent upon the vendor.
|
| 14 |
See Patent No. 4,675,863, June 23, 1987.
|
| 15 |
See McGarty, Columbia Paper, March, 1996, Disaggregation. In that paper the author develops the theory of disaggregation. Also see McGarty, Federal Communications Law Journal, in which this theory is extended.
"What the theory states is simply: The theory of disaggregation states that technology and industry has developed in such a fashion that it is possible to effect all elements of a business in a virtual form by obtaining all functions necessary to del
iver a service by purchasing them from third parties each of whom has themselves other similar customers and thus each of whom can deliver their element of the functionality in a minimal marginal cost manner. The disaggregation theory then concludes with
the result that in many technologically intense services business, a virtual company can exist wherein all the functions can be purchased from third parties or capital equipment may be purchased in a fully interconnected fashion so as to achieve near equa
lity between average and marginal costs from the very commencement of the business. The Disaggregated Company is the embodiment of the virtual business."
|
| 16 |
The calculation for analog goes as follows. each analog signal occupies 30 KHz of Bandwidth. There are 42 such channels in 1,250 KHz. Yet there is a reuse factor of 7 for analog. This means that for every cell, the surrounding 6 cells cannot use the same
frequency amongst them, thus the 42 number must be divided by 6, yielding 7 channels. The CDMA system thus has a 10:1 ratio over analog.
|
| 17 |
It should be noted that in the C Band auctions the bidders bid an average of $50 per PoP. The F Band is less than one tenth that number. However, from the above analysis, there is less than an 15% capital penalty for the use of 10 MHz CDMA. Thus the lower
bid price must reflect the dilution of the market from the larger number.
|
| 18 |
See the papers by McGarty in Interconnect and access. These discuss the detailed economics and the costs models. Also see the paper by McGarty, Federal Communications Law Journal. This latter paper presents a detailed antitrust analysis of the interconnec
tion issue.
|
| 19 |
See FCC GN Docket No. 93-252, February 3, 1994, FCC 94-31; ¶ 37.
|
| 20 |
See Title 47 C.F.R . §51.701(b)(2)
|
| 21 |
Filed with Massachusetts DPU on August 22, 1996.
|