CC Docket No. 96-45
REPLY COMMENTS OF PACIFIC TELESIS GROUP
Pacific Telesis Group files these reply comments in response to the Notice of Proposed Rulemaking ("NPRM") released March 8, 1996. Hundreds of parties filed comments in this proceeding; we will address some of the concerns and arguments raised by some of the parties.
HIGH COST FUND
Services to Be Included in Definition of Universal Service
Some parties argue that ISDN and other "advanced" services should be included in the definition of basic service. While we believe in promoting these types of services, we agree with Nynex and others that they have not risen to a level which justifies their inclusion as
1 See, for example, Alaska PUC, pp. 1-6; Alaska Telephone Ass'n, pp. 2-3; Cheyenne River Sioux Telephone Authority and Goldenwest Telecommunications Co-op, p. 7; Governor of Guam, p. 7; Matanuska Telephone Ass'n, p.2; Montana Telephone Ass'n, p.2; Western Alliance, p. 12.
2 See, for example, NYNEX, p. 12; AdHoc Telecommunications Users Committee,
pp. 4-5; ACTA, pp. i, 3; BellSouth, p. 23; California PUC, pp. 4-5; General
Communications, Inc., pp. 6-
AT&T and TCA, Inc. propose that local number portability be included in the list of core services to be covered by the universal service fund. Section 25 1 (e)(2) provides that the costs of establishing number portability should be "borne by all telecommunications carriers on a competitively neutral basis as determined by the Commission." The universal service funding mechanism would meet this requirement. We believe that recovery of the costs of number portability in this manner could be accomplished by folding those costs into the universal service fund. However, the disbursement of dollars collected for number portability costs must be directed to the party incurring the costs (as opposed to the party serving that particular customer).
7; Georgia PSC, p. 6; GTE, pp. 2-3; MCI, pp. 22-23; New Mexico Attorney General, p.4; Fred 'Williamson and Associates, p..7.
3 See, for example, Georgia PSC, p. 10; GTE, pp. 2-3; NYNEX, p. 12; Southwestern Bell Telephone Company, pp. 8-9; Tele-Communications, Inc. pp. 7-9; See also Remarks of Commissioner Chong at Joint Board Meeting April 12, 1996.
4 Currently, we have about 70,000 customer lines of ISDN, out of a total of about 15 million access lines.
5 AT&T, pp. 12-13; TCA, Inc., p. 5.
-- 2 --
AT&T, MFS, and NASUCA argue, without support, that retail local service rates already recover efficiently incurred Costs. This is dead wrong. As AT&T knows,  the cost of local service traditionally has been higher than local phone rates, and subsidized by high toll rates. Our model, and the Benchmark Cost Model ("BCM"), both show many areas where local service is below cost. Prior California Public Utilities Commission ("CPUC") decisions however, have accepted Pacific's cost studies and have acknowledged that the price for basic residential service is below direct embedded cost. Moreover, AT&T goes on to argue that while local service doesn't need a subsidy, interexchange carriers should be able to dip into the subsidy fund to recover costs for interexchange traffic provided below cost to low-income consumers or for calls to or from high cost areas. But, the cost of local service has far greater geographic differences than the cost of long distance service.
Furthermore, NASUCA erroneously claims that Pacific Bell's cost studies "grossly overstate the cost of basic exchange service." These cost studies, currently being reviewed by the CPUC, proceed from an industry-negotiated set of principles for costing the Total Service Long Run Incremental Cost ("TSLRIC.") The studies adopt a "scorched node," long-run view, with forward looking technology assumptions, and size the technology according
6 AT&T, p. 2; MFS, p.5; NASUCA, p. 12.
7 If rates were higher than costs, facilities-based residential competition would be starting. Instead, in California, AT&T shows every intention pursuing a resale strategy to compete for the residential market since it knows that it cannot provide service on a facilities basis at current rates and remain profitable.
8 D.94-09-065, mimeo, p. 40 "We continue to price residential basic access service below cost to mitigate the effect of higher total monthly bills on customers with low toll usage."
9 AT&T p. 12.
10 NASUCA, p. 20.
-- 3 --
The studies contain the most extensive level of geographic deaveraging yet achieved. This is a long-missing feature of telephony cost studies, which as been a problem since costs tend to be quite sensitive to geographic variation in density, loop length and terrain, and competition will key on these cost differentials. The studies are being intensely reviewed by the industry and the CPUC.
We maintain our continued commitment to affordable basic rates. We believe that deaveraging the subsidy will help maintain these rates.
Use Of Our Cost Proxy Model (CPM)
Many parties supported the use of a proxy model to establish costs of serving customers on a geographically deaveraged basis. A proxy model allows costs to be disaggregated so that costs of service can be figured on a geographic basis (such as census block group or 1/4 square mile grid cell). Our proxy model is a model designed for use by regulators. All inputs and algorithms can be adjusted. It can accommodate inputs based on forward looking costs, historical costs, embedded costs, actual costs, costs determined based on proxy factors, or a combination of any of these. Commenters who, objected to the CPM based on its use of
11 See, for example, AT&T, Appendix A ages 1-2; State of California and California PUC, pp. 9-1 0; Florida PSC, pp. 9-11; LDDS WorldCom, Inc., p. 12; MFS Communications Co., p. 18; Nat'l Cable Television Ass'n, Inc., pp. 7-8; Tele-Communications, Inc., p. 11.
-- 4 --
proprietary costs are misinformed. The CPM can now use either commercially available customer input data or company-specific customer location data. For rural areas, the commercially available data is at a finer level of detail than census block group data, ensuring more accurate costing for those areas.
A diskette containing the CPM populated with a sample data set of publicly obtained data from California is available for companies who wish to further examine the CPM. A user's manual is also available. Contact undersigned counsel for further details and licensing arrangements to review the model.
A cost proxy model will allow the Commission to carry out the terms of the Act in an efficient and deregulatory manner. There is no consensus in the comments as to whether a proxy cost model or actual, embedded costs should be used. A proxy model is a way to simulate the actual costs involved in providing service on a disaggregated level. If another method exists for geographically deaveraging actual costs, we would be happy to use that method, but for now, we believe a proxy cost model is necessary for this calculation.
The important step in calculating the universal support obligation is to determine the true costs, whether from a model, or from embedded costs. A model has advantages of administrative simplicity and consistency, but is dependent on the inputs used. Actual costs will accurately represent the costs incurred by the carrier, but may differ widely among carriers, are difficult to geographically deaverage, and may not encourage efficient provision of basic service.
The proxy model simplifies administration in that once an appropriate model is developed, and the appropriate inputs are agreed to, the model runs the same way for each carrier or class of carrier. There is no need to review each carrier's actual costs and compare them to
-- 5 --
Whatever model the Commission endorses, the model must have the following attributes: 1) it must accurately include all network elements; 2) it must be based on the most modem technology currently being deployed, not on futuristic, wireless technologies as MCI suggests; 3) it must recognize efficiencies and differences between large and small carriers so that appropriate costs are included; 4) it must model a realistic distribution of population; 5) it must include a reasonable amount of shared and common costs; and 6) it must be verifiable.
The model must accurately represent the existing technology. A model based on an imagined network of the future will not allow current companies to recover the legitimate costs of serving customers. If new, lower-cost technologies arrive on the scene, they should be encouraged to be deployed where they are most efficient--where they can make the greatest efficiency improvement over the existing technology. Once the new technology is in wide use as a universal service alternative, then the Commission can revise the inputs to the model and adjust the subsidy amount using data from actual experience.
Many small rural carriers surmise that a model will not allow them to recover appropriate amounts for overheads and other costs particular to small companies. However, our CPM, by allowing for different inputs, can accommodate small carriers by handling flexibly any requirements for a change in the investments, overheads or other costs. The Commission could order, for example, that overheads for small carriers be computed based on a different
12 MCI, P. 11.
13 Rural Telephone Coalition, p. 16; Frederick & Warinner, p. 3.
-- 6 --
input from large carriers so that no carrier's costs would be incorrectly estimated. Similarly, it could order that the current system of DEM weighting for small carriers be imported into the model.
We agree with commenters that shared and common costs must be included in the costs calculated to determine the universal service support level. Without some recovery of shared and common costs, real costs of serving a customer will not be recovered. Also, failing to include shared and common costs is contrary to the Act's directive to encourage competition, and to explicitly fund universal service. If TSLRIC were used to set the cost of universal service, with no contribution to the joint and common costs of the enterprise, all other services provided by that carrier would need to carry a disproportionate share of joint and common costs. The effect would be to artificially raise the costs and the prices for those other services, precisely at the time the Commission is charged with opening the markets to competitive forces.
Similarly, NASUCA claims that the basic service rate element must not bear the total cost of the loop. This position has been rejected by the mainstream economic literature years ago. Loop cost are incremental costs of basic access service. They are avoided if, and only if, the firm discontinues basic access service. They are not attributable to any other service. Customers who make no long distance calls, or whose lines have toll restrictions on them, still incur full loop costs. Therefore, it is appropriate for the total cost of the loop to be included in the calculation of basic service costs.
14 See, for example, Montana Telephone Ass'n, p. 7; Cf.,AT&T, p.7. MCI, p. 5, advocating TSLRIC to determine the costs to be supported by a universal service fund.
15 NASUCA p.21.
16 Alfred E. Kahn and William B. Shrew, Current Issues in Telecommunications Regulation: Pricing, 4 Yale J. on Reg. 191 (1987).
-- 7 --
Any model must take into account geographic differences in the network costs, and must model a realistic population distribution. The Commission should avoid averaging of geographic differences to the greatest extent possible. Many rural census block groups contain significant areas of population clustering. Thus, costs in the census block group vary greatly depending on the distance of the customer from the central office. Many census block groups may also be served by more than one carrier. Any model adopted must take these attributes into account.
Finally, the model must be verifiable so that reasonable audits can be done to assure regulators that true costs are being identified. While many commenters identified the proprietary nature of the CPM as a reason to disregard it, as we said in our comments, the model has a version which uses commercially-available customer data, so that all inputs can be fully verified by other parties.
Some carriers support the concept of proxy models but note that such models must be rigorously examined, tested and evaluated. We do not disagree with these statements. However, the Commission and Joint Board do not need to perform that work themselves. That testing and evaluation are going on in the states, particularly California. The CPUC is currently holding evidentiary hearings on the Hatfield Extension to the Benchmark Cost Model and our CPM. The CPUC expects to adopt one of these models, issue an order by August of this year, and have the Universal Service Fund operational and funded by October. The Joint Board and the Commission can utilize the work going on in California so that the model chosen has the
17 See, for example, Citizen's Utilities Company, p. 13; Joint Filing of Maine PUC and 8 other State Commissions, p. 5; Minnesota Telephone Association, p. 1.
-- 8 --
An alternative approach, suggested by U.S. West is for the Commission to convene an industry task force to come up with a model which meets the criteria it specified. Because so much work has already been done in this area, such a task force could be given a very short timetable, perhaps 60 days to come up with a resolution. That model would then be used to determine the proxy costs.
Calculation of the Subsidy
The subsidy amount must be calculated properly so that appropriate incentives for competitive entry and for providing service exist in the market. We do not agree with those commenters who argue that the subsidy amount should be the difference between the costs of serving the area and the average cost of service. The average cost of service has no relevance to the revenue needed to sustain universal service. Calculating the subsidy on this basis will not make the subsidy explicit, as required by the Act. If the difference between cost of serving the area and average cost of service is used as the subsidy determinant, it will not allow a carrier to recover its costs in areas that are truly high cost (but don't fall above the average). Subsidy should be calculated based on the difference between the estimated costs to serve and the price charged. the Benchmark rate (set at a level which uses the available funds in the federal universal service fund, currently about $5 Billion attributable to Carrier Common Line and USF) can be used to allocate the subsidy between the federal plan and the corresponding state plan.
18 47 U.S.C. [[section]]254(b)(4).
19 U.S. West, p. 10.
20 MCI, p.4; Telecommunications Resellers Association, pp. 12-14.
-- 9 --
The subsidy should go to the carrier incurring the cost but not receiving the revenue. For example, if a reseller pays the full deaveraged cost (including some recovery of shared and common costs) and that cost is above the benchmark, then the reseller gets the subsidy. On the other hand, if a reseller purchases a line at rates below the full deaveraged cost, then the facilities-based carrier should receive the subsidy. Contrary to LDDS Worldcom's argument,  a reseller who buys the service with no use of its own facilities should not be entitled to receive a subsidy, unless the reseller has purchased those facilities at the full deaveraged cost of those facilities. Otherwise, the reseller has the incentive to serve only areas where the average price most exceeds the universal service subsidy, thus collecting a windfall. The facilities-based provider then loses the subsidy but retains the high cost of serving the customer.
Another issue raised by various commenters (mostly enhanced service providers, or Internet access providers) is whether they meet the definition of a carrier for purposes of contributing to the fund. The Commission should require that, to the extent an entity provides telecommunications services, (irrespective of what other services they might provide as well) they are telecommunications common carriers under the Act, and should contribute to the universal service fund. See 47 USC [[section]]153(44). The Commission may need to determine, on a case-by-case basis whether a particular entity is in fact providing telecommunications service. By undertaking such a review the Commission can avoid giving entities the incentive to dodge universal service support by bundling its telecommunications service with information service.
We agree with Comptel and Telecommunications Resellers Association that another component that should be added to the subsidy fund is that portion of the Residual
21 LDDS Worldcom, p. 5.
-- 10 --
We also agree with SBC and others who support recovery for underdepreciation of equipment due to regulatory control over depreciation rates. Such amounts are holdovers from monopoly regulation and the inclusion of them in LECs' rates makes it extremely difficult to compete with new providers not so burdened. Recovering these amounts on a competitively neutral basis will allow an even playing field to be created for competition in telecommunications. We agree with SBC that a separate and explicit fund should be set up to recover these amounts.
Services to be Supported
There is no consensus about what services should be supported for educational purposes, or at what level the discount for those services should be set. Understandably, schools want a variety of advanced services available to the school at a very affordable price, in some cases, free. While this might be a laudable long-term goal, certain guidelines must be put in place to ensure implementation of such long-term objectives. We continue our long-standing support of reasonable and necessary service discounts to schools and libraries (we already do this in California under our Knowledge Network Rate for ISDN, our California Research. and
22 Comptel, p. 15; Telecommunications Resellers Association, p. 14.
23 SBC, pp. 23-24.
-- 11 --
The Commission and the Joint Board must also consider the public's willingness to pay for the educational services given at a discount. The Act requires every telecommunications carrier to contribute to the universal service fund. Presumably, these carriers will pass along these costs to consumers. Because the subsidy fund relating to education and health care is a new subsidy fund (as opposed to the high cost fund, which is simply taking an implicit subsidy and making it explicit), all dollars earmarked for this fund are additional dollars that consumers will be paying. Thus, the support for educational access to technology must be balanced against consumers' willingness to pay. We agree that many of the proposals are interesting and, if there were unlimited funds, would be worthwhile. However, the Commission must consider how these advanced services will be funded.
Many of the proposals championed by educational institutions commenting in this proceeding advocate far more bandwidth and speed than may be necessary or appropriate given that there are many schools and libraries which lack even rudimentary access to the information superhighway. For example, some commenters seek to include T-1 access, ATM and other advanced services as part of the supported universal service. Less expensive, but equally valuable technologies, however, can provide education institutions with the necessary access to distance learning, videoconferencing and Internet access. T-1 will serve about 250 individual computer terminals,  but requires sophisticated and expensive electronics on site at the school.
24 Section 254(d).
25 Ten times the average number of students per classroom and twenty times the associated number of computers.
-- 12 --
Thus, even if T-1 service were provided for free, few schools could afford to buy the equipment needed to make use of it. An ISDN line, on the other hand, easily serves about 30-40 computers simultaneously. Five ISDN lines, as our Education First program provides, usually support the needed number of computers on site at a far cheaper cost, and facilitate most videoconferencing and Internet access application requirements.
We agree with those parties, though, who advocate a standard based on capabilities, rather than a technology-specific criteria. By stating what capabilities need to be supported (Internet access, for example), the schools, and the telecommunications carriers serving the school, can determine the best way to achieve that standard.
Allocating the Federal Education Fund
Many parties have discussed how to design and disburse a federal education fund in a manner which is equitable and which will accomplish the goals of the Act. The Senators who sponsored the amendment which created section 254(h) recently wrote to the FCC stressing the intent of the amendment in fostering affordable access to all schools. In order to ensure that all schools are brought up to benchmark level of technological deployment, we support disbursing at least a portion of the federal fund on an "as needed" basis. We suggest that funds be disbursed on a 50/50 basis, half on a per capita or per student basis and half on objective, technology-based criteria. For example, one noted criterion is the number of students per computer per school. (This sort of data is available through third party vendors and is compiled
26 This approach is supported by the National Information Infrastructure reports, as well as the California SB600 Task Force Report (which studied telecommunications infrastructure in' California K- 1 2 schools and public libraries).
27 Letter from United States Senators Olympia J. Snowe, John D. Rockefeller, IV, J. James Exon and J. Robert Kerrey to Reed Hundt, April 24, 1996, at pp. 1-3.
-- 13 --
We agree with NYNEX 29 who suggests that an Educational Advisory Council should be created (similar to the Telemedicine Advisory Committee recently announced). The Education Council could be charged with reviewing the benchmark set by the Commission and the Joint Board on a periodic basis to ensure that an appropriate level of technology deployment is being encouraged.
State Involvement/Creation of Benchmark
We agree with many parties who support state involvement in the education fund as the most helpful and efficient way to implement the educational access to technology provisions of the Act. 30 The State of California has already taken action in this area. The California Education Technology Task Force, commissioned by the State Superintendent of Schools, sets out a process for educational technology to be deployed to California schools. It advocates a state benchmark for technology deployment in schools. These benchmarks do not specify technology type but contain minimum standards with which each school must be equipped. Each district must submit a plan for how it will meet the benchmark. This approach
28 This figure is often quoted as the average level of technology deployment today.
29 NYNEX, pp. 20-22.
30 BellSouth, p. 221; New Jersey Division of Ratepayer Advocate, p. 23; United States Telephone Association, pp. i, 6; Wisconsin PSC, p. 17.
-- 14 --
Size of the Fund
The Task Force concluded that $10.9 Billion over 4 years (or a current revenue requirement of $7.5 Billion) would be necessary to fund the chosen technology benchmark. That amount was calculated by reviewing the total amounts necessary for a successful technology deployment. The breakdown in the Report is as follows:
Staff development and Support 21% of total
Courseware, Materials and Services
(including communications services) 26% of total
Hardware and Telecommunications
Infrastructure 53% of total
The report quantifies the dollars needed to fund all of the different aspects of deploying technology in schools. The Report concludes that amounts related to telecommunications are approximately $112M for communication charges. Because the Report deals only with public K-12 schools, that number needs to be adjusted upwards for private schools and libraries. When those institutions are added, the amount needed for telecommunications services is approximately $166M over 4 years.
The Report also calculates approximately $547M for telecommunications infrastructure for K-12 public schools. That amount is principally attributable to the cost of inside wiring. Inside wiring has been deregulated by the Commission and should not be part of a mandated fund. Moreover, given the Commission's recent NPRM on unregulated wireless
-- 15 --
Thus, in California, $166M over 4 years, will support deploying access to telecommunications services for the approximately 12,000 schools and libraries. The level of services supported by this amount is $300 per school per month (for 4 years) for telecommunications services. In California, this equates to about 5 ISDN lines. Five ISDN lines can support about 150 computers within the school as well as compressed video for distance learning applications. We support this level of technology deployment in the schools.
To figure the size of the Education Fund, a discount rate needs to be calculated (the $166M over 4 years is the total cost). At a 25% discount, the Fund would need to supply $42M over 4 years (approximately $IOM per year). At a 50% discount, the Fund would need to supply $83M over 4 years (approximately $2lM per year). Of course these numbers are just for California, and would need to be extrapolated nationwide for the Federal Fund. 
We have closely worked with this Task Force, and heartily endorse its conclusions and approaches. Since this work has already been done at the state level, it seems
31 "Commission Proposes to Make Spectrum Available for Use by New Licensed Equipment; NIISUPERNet Devices would facilitate Interconnection to Nll," Report No. 96-36, Action in Docket Case, April 25, 1996.
32 These figures only include the cost to fund the telecommunications charges for advanced services. These figures also do not include a discount on basic service (e.g. I MB lines') for administrative use by the school.
33 The final Report of this Task Force will be issued shortly. We will submit it to the Commission and the Joint Board when it is available.
-- 16 --
redundant for the Commission to establish its own processes for achieving the same goal. Thus the Commission could establish minimum national standards to ensure some level of deployment of technology, and then allow the states to determine what best meets the needs of the educational community in that state. 
ADMINISTRATION OF THE FUND
Calculation Of The Surcharge
A single surcharge large enough to cover the high cost fund and the education and health care funds should be assessed in a unified manner. The administrator should then disburse the funds in accordance with Commission directives.
Our comments contain our recommendation that the surcharge required for the universal service fund(s) be collected based on a carrier's interstate revenues. After review of the various positions, we agree with commenters who suggest that the surcharge should be calculated on the interstate revenues less payments made to other carriers.  This will enable carriers to avoid paying the surcharge on amounts paid to other carriers. We further suggest that the surcharge not be paid on the SLC since the SLC already recovers costs relating to the loop. Putting a surcharge on the SLC directly increases the price and thus the cost difference and so increases the subsidy need.
34 In Nevada, we have filed a proposal to create a fund to be used to provide educational access , to technology through dollars earmarked for sharing with consumers. Nevada Administrative Code [[section]]704.6843. 1. This proposal is currently under consideration by the Public Service Commission of Nevada.
35 See, for example, LDDS WorldCom, p. 19.
-- 17 --
Administration of the High Cost Fund
As to the actual fund administrator, we agree with many parties that NECA would be a good choice to administer the federal high cost fund in that it has the expertise in collecting and disbursing funds of this type , and met the requirement of being a neutral third party. The Education and Health care funds should be administered on a state basis in order to target funds where they are most needed. NECA could collect the surcharge for these funds, and then disburse them to the states in accordance with Commission guidelines.
LOW INCOME ISSUE
The National Association of State Utility Consumer Advocates, the Ohio Consumer Counsel & the Benton Foundation advocate a prohibition on disconnection for nonpayment of toll charges. 39 We do not support this mandated approach. We believe that each
36 Sprint, p. 16.
37 If the surcharge were on both interstate and intrastate revenues, AT&T would be able to dilute paying its- fair share based on its substantial interstate revenues.
38 See, for example, ATA, p. 9; Rural Telephone Coalition, p. 1-2; Shawnee Telephone Company, p. 1; South Carolina PSC, p.2; Southwest Bell, p. 20, n.33; TCA, p. 7; Western Alliance, p. 13; Wyoming Public Service Commission, p. 5.
39 NASUCA, p. 6, Ohio Consumer Counsel, p. 16, Benton Foundation, p. 2.
-- 18 --
We believe that a product offering of toll restricted service enables customers to remain connected while paying off an outstanding balance. In this scenario the customers retains ultimate responsibility for payment. and the root cause of the problem, control of telephone usage, is addressed. We are confident that our toll restricted service (as outlined in our original comments) will positively impact subscribership rates in the long term. With appropriate education e.g. community outreach programs, customers can take advantage of toll restricted service and retain access to the telecommunications network.
In states where disconnection has been prohibited, net bad debt has increased significantly (as reported by Pennsylvania Telephone Association, an increase of 300% in net bad debt). These losses are ultimately shared by the entire customer base. Customer confusion regarding payment options and implementation of multiple billing systems are other implications of such a prohibition.
The Commission should adopt rules which allow high cost areas of the country to be supported by the universal service fund on a census block group (or smaller) basis. The cost of providing service should either be calculated on an actual cost basis, or by use of a proxy model, or some combination of the two. A competitively neutral surcharge on revenues should
40 Federal State Joint Board Open Meeting, April 12, 1996.
-- 19 --
The Education Fund should be disbursed to the states on the basis of a dollar-per student benchmark, combined with a benchmark based on need for the technology (such as number of students per computer). The states can then apportion the funds to school districts needing the money based on the district's plan for attaining appropriate benchmark levels. We
-- 20 --
PACIFIC TELESIS GROUP
RANDALL E. CAPE
NANCY C. WOOLF
140 New Montgomery Street, Rm. 1523
San Francisco, California 94105
MARGARET E. GARBER
1275 Pennsylvania Avenue, NW
Washington. DC 20004
Date: May 7,1996
-- 21 --
CERTIFICATE OF SERVICE
I, Colin Petheram, hereby certify that on this 7th day of May, 1996 a true and correct copy of the foregoing Reply Comments of Pacific Telesis Group was mailed, first class-postage prepaid, to the parties shown on the attached list.