[2] See, for example, comments of the ICORE companies.

[3] GTE defines the term COLR to mean an Eligible Telecommunications Company ("Eltel") as defined under the '96 Act that undertakes obligations established by a state agency, within Federal guidelines, as a condition for receipt of Federal universal service support.

[4] This rules out the false solutions -- presented by such parties as AARP at i-ii -- that essentially call for continuation of the current system. The statute dictates a new system meeting new criteria.

[5] See Ameritech at 5; AT&T at ii, 21; ALTS at i; GTE at 16-18; CompTel at ii.

[6] Florida Cable Telecommunication Association at 8. See also, e.g., MFS at 4; LDDS WorldCom at 2-3; CompTel at 9; Time Warner at 5; AirTouch at 5.

[7] Congress mandated that the Joint Board/FCC fashion a program that is "specific, predictable and sufficient ... to preserve and advance universal service." [[section]]254(b)(5). New programs must connect the payment of universal service support to actions and offerings of local service providers that will promote universal service.

[8] As the CPUC notes at 8, it may be reasonable to expect rural, high cost customers to pay more than urban customers, within some reasonable limits. However, where a binding rate constraint is necessary, it should be applied symmetrically to all carriers that receive funding. The CPUC at 3 stresses the importance of this neutrality, and says at 8 that it plans to base support on the difference between the cost of providing service and "the rate that local exchange carriers will be authorized to charge."

[9] See also: "The OPUC has determined that service providers must meet minimum service quality standards in order to qualify for universal service support funds.... The service quality standards set by the OPUC are not a requirement for market entry; they are a requirement for obtaining universal service support." Emphasis added. Oregon Public Utility Commission ("OPUC") at 2-3. "The VSCC required new entrants to comply with the service quality criteria that have been applied to incumbent LECs for many years.... We urge the FCC to base its quality measurements on existing standards in the states." Virginia State Corporation Commission Staff ("VSCC") at 1. The VSCC adds at 6: "We have long held in Virginia that poor service at any price is no bargain. Affordable rates must always be based on good service." The Public Utility Commission of Texas at 2-5 expresses special concern with maintaining levels of quality.

[10] See Ameritech's discussion of unilateral requirements as opposed to bilateral requirements. Ameritech at 5 and n.6 and n.7.

[11] The Communications Workers of America at 7-9 discussed the quality of service provided by GTE. Their concerns were apparently shaped by the data available in the Commission's ARMIS 43-05 reports. GTE had already notified the Commission that it had discovered an error in the automated system that generates data for the ARMIS reports. This error caused an overstatement of repair measurements during the reporting period (4Q1994 through 3Q1995). See letter of Edwin J. Shimizu, Director-Regulatory Matters, to Mr. William F. Caton, Acting Secretary, dated April 12, 1986.

[12] GTE recognizes that a reasonable transitional period may be necessary to implement this program. See USTA at 18, NARUC at 11, OPUC at 6, Florida Public Service Commission at 12, WUTC at 11-12, NECA at 14, and the Missouri Public Service Commission at 8.

[13] See also Sprint at 8: "It is critical that only services that a majority of consumers have subscribed to be funded because ultimately it is the consumer that funds the subsidy."

[14] Emphasis added. While WUTC warns against regulatory action picking technological winners and losers, MCI at 16 would plunge regulation into precisely that morass, since MCI would have technology specified as part of the plan: "[T]he LECs should be required to use all digital switches and provide SS7 signaling throughout their network." As correctly stressed by the wireless carriers, the definition of core service should not be linked to any particular technology. See Western Wireless at 7-9, AirTouch at 10, and PCIA at 14-16.

[15] These features are used by a "substantial majority" of residential customers, but not all features are ubiquitously available today. A transition plan will be needed to meet these standards nationwide. For example, locations where the transition to single party service has not yet been completed should continue to receive support for party line service.

[16] See, for example, NCTA at 6 and Attachment A at vi; Illinois Commerce Commission ("Illinois CC") at 5; Ad Hoc at 12. NCTA's Attachment A at 101-106 develops an elaborate argument that ascribes virtually all of the costs of any spare capacity in the network to second lines. In fact, any network is engineered to provide a combination of services. Spare capacity simply represents the engineer's trade offs between the frequency and magnitude of network additions, given that it is usually very costly to add actual distribution plant. Factors that affect the achieved level of utilization in a network include growth in demand, indivisibilities of basic network resources (such as minimum cable sizes), uncertainty concerning demand (based perhaps on competition for basic service), and the need to provide additional lines. The engineer cannot distinguish between two lines provided to a single family dwelling, and two lines provided to a dwelling that is shared by two households.

[17] A broadly expressed concern is to avoid "haves" and "have-nots" with respect to information services and access to the Internet. Access to most such services is possible today, using "core" service and a computer equipped with a modem. However, any prolonged access to such services using a single line would preclude the use of the line for other purposes -- such as receiving calls in emergency situations. Parties proposing that the definition should be limited to a single line simply assume -- without any supporting argument, or evidence of Congressional intent -- that the national policy objectives will be fully satisfied by the provision of only a single line per unit.

[18] In recent testimony before the CPUC, Dr. Nina Cornell, appearing for AT&T and MCI, suggested that one line be provided to each "household." When asked how a household should be defined, she suggested such criteria as whether a person commonly took meals with other household members. Dr. Lee Selwyn, also appearing for AT&T and MCI, proposed a different criterion, in which one line would be provided to a "dwelling." He suggested that a COLR, in responding to a customer request for service, should consult municipal zoning records to determine if the customer occupied a separate "dwelling." In cases where a room or other portion of a structure were rented, he suggested that the COLR should inspect the rental agreement. Any of these suggestions would be not only burdensome but unworkable for the serving COLR.

[19] Indeed, households with modest incomes, or who have recently moved to a new area, are probably the most likely to be living in some shared arrangement that would make it difficult to distinguish one household from another. These are also the households for whom concerns over the need for affordable service are the greatest.

[20] The NPRM seeks comment at [[paragraph]]24 as to whether the Federal plan should be designed to achieve an objective rate. A wide range of parties joined GTE in proposing that the Federal plan should establish rate thresholds based on policy determinations as to what rate level is affordable. See Michigan Department of Commerce at 2; Illinois CC at 7; MFS at 18; General Communication at 9; Time Warner at 7; US WEST at 8, Ad Hoc at 17-18, MCI at 4, Sprint at 4, 9, AT&T at 14 and Appendix B.

[21] States that restrict the rate the COLR may charge to a level less than the affordable level should fund the entire difference between the permitted rate and the cost, not just the difference between the threshold and the cost. The Federal plan could include a guideline which requires the state plan to be "sufficient" in this sense, as a condition for the provision of Federal funding toward offsetting reductions in state rates. See also AT&T at 15.

[22] The separations treatment of the support can then be adjusted to match the policy judgment made by the FCC in setting the thresholds. Such an adjustment for LECs would be similar to the way USF works today. However, separations should accommodate the FCC's policy, not drive it. See, for example, US WEST at 13.

[23] It will be essential that the Joint Board/FCC plan approach with great skepticism any claim to be entitled to receive support and/or be eligible for reciprocal compensation under [[section]]251 combined with a denial of obligation to pay the appropriate contribution to the universal service fund. See Telecommunications Resellers Association at 6 and 9.

[24] See AT&T at 9: The use of surcharge on both interstate and intrastate retail services "obviates altogether the potentially difficult problems associated with having to make jurisdictional determinations."

[25] AT&T at 8 says: "A surcharge on all retail telecommunications services, both interstate and intrastate, creates a fair, simple and efficient recovery mechanism." Footnote omitted. CompTel at 15 recognizes that "a tax assessed on end user retail revenues" would be the "most equitable way to collect universal service funds...."

[26] The current TRS method does not satisfy this requirement because ILEC access revenues are counted twice: once as ILEC access revenues, and a second time through the price for interexchange service that must be set to recover those input costs. While the resulting distortion may be tolerable in a relatively small fund, such as TRS, it would certainly present a major problem in a mechanism large enough to support the requirements of the '96 Act. NPRM at [[paragraph]]122.

[27] The CPUC already employs what it calls an "All End User Surcharge" to fund its state lifeline program, which requires more than $300 million per year. Vermont has also implemented a retail surcharge to fund its universal service program.

[28] Using a retail surcharge, the carriers are simply the "tax collectors" for the fund. Issues raised by some parties as to which carrier should "contribute" more or less are thus moot. The retail surcharge approach recognizes that all revenues come ultimately from customers.

[29] Proposals that ILECs meeting the criteria of the Joint Board/FCC plan but falling within certain arbitrary regulatory classifications be ineligible for universal service support are clearly in conflict with the '96 Act. See, e.g., Time Warner at 11-12 (only rate-of-return ILECs should be eligible for support) and NCTA at 13 ("must seriously consider whether support is required for price cap LECs").

[30] Under the bidding approach, the sum of the required COLR rate and the support determined through the auction process will be an estimate of the market rate.

[31] See, e.g., AT&T at 6-7; MCI at 4; TCG at 7.

[32] TSLRIC studies also may be useful to regulators in determining if cross-subsidies exist. TSLRIC is not useful in establishing a cost floor for individual rates. Instead, it establishes a floor for the revenue that all rates for the service must generate. This will be equivalent to a price floor only in the special case where there is a single, uniform rate for all units sold.

[33] There is a fourth possible cost -- a "residual" cost. A regulated multi-product firm may have residual costs caused by assets that remain on its books through the actions of regulatory requirements even though they have no current economic value. These assets remain as a cost even though they cannot produce any positive cash flow.

[34] This approach is also consistent with the FCC's current rules for the pricing of new access services, which is based on direct cost plus a uniform overhead loading.

[35] NCTA suggests at 10 that the use of CBGs will somehow "fail to recognize the economies of scale of serving several CBGs from a single wire center...." NCTA therefore proposes that costs be estimated at the wire center level. NCTA fails to understand the design of existing proxy models. Both the Benchmark Cost Model ("BCM") and the Pacific Bell model ("CPM") engineer plant at the wire center level for all of the CBGs assigned to the wire center. They therefore capture the economies NCTA is concerned about.

[36] NPRM at [[paragraph]]34. Many parties agree on the use of small units such as CBGs. See, for example, CPUC at 9 , NYNEX at 10, US WEST at 8, Sprint at 15-16, MCI at 10-11.

[37] Use of a small geographic area makes academic consideration of changes in the definition of study area. NPRM at [[paragraph]]45.

[38] See, for example, AT&T at Appendix B; NCTA, Attachment A at Appendix 8B.

[39] Within a limited number of zones, the actual density of each CBG could vary widely. Further, every existing proxy model considers explanatory variables other than density. Variations in cost attributable to these factors will not be captured when CBGs are grouped according to density. Within each zone, CBGs with lower costs and/or higher revenues will offset CBGs with higher costs and/or lower revenue. The effect of CBG aggregation would be to underestimate the true funding requirement, and to target support inaccurately.

[40] The Illinois CC at 6 explains that any attempt to make a model technology-neutral would cause it to lose its ability to estimate meaningful costs. The CPUC at 12 also recommends starting with "a wireline model" because this approach would "encourage the carrier with the least-cost technology to offer service, without prejudging which technology should be used."

[41] While the CPUC at 12 says that market conditions may not warrant the introduction at present of a competitive bidding process to determine high-cost support amounts, it does not foreclose possible support for a bidding procedure in the future. Indeed, it says (id.) it is "considering using competitive bidding to determine subsidy amounts once they come up for review" because of "the level of difficulty experienced in getting parties to agree on a proxy cost model."

[42] Time Warner at 10 rejects a bidding procedure that would award the lowest bidder exclusive high cost support for a particular area, but this does not describe GTE's plan. It adds at 11: "Absent such an alternative [a non-preclusive bidding procedure like GTE's proposal where bidders are provided "with the greatest incentive to bid efficiently, ensuring that support would be provided at the least cost"], it is imperative that the Commission implement a bidding mechanism that will provide a similar incentive."

[43] See (i) Appendix C to GTE's comments for a description of GTE's proposed bidding process submitted in D.80-286 and (ii) GTE's comments at 10-12. The NPRM at [[paragraph]]36 proposes that a winner's preference, or "incentive bonus," could be provided to the lowest bidder in an auction as an incentive to bid aggressively. GTE has proposed a process wherein the size of this incentive bonus is itself determined through the bidding process.

[44] For LECs that receive USF today, the offsetting rate reductions should be based on the net change in support caused by the new plan. For non-LEC Eltels, no offsetting reductions will be required, since these carriers do not provide implicit support through their rates today.

[45] The CPUC says at 20 it recommends elimination of the CCL, which is "not explicit because it is embedded in [LEC] access rates." As the CCL is eliminated, perhaps through some transition period, the need for Long Term Support to equalize CCL rates for NECA pool companies will also be eliminated. NPRM at [[paragraph]]115.

[46] Time Warner at 20; PCIA at 13; CompuServe at 6-7; MFS at 22.

[47] As discussed in GTE's comments at 12-14, GTE shares the concern stated in the NPRM at [[paragraph]]30 that the current Part 36 separations process does not provide a sound basis on which to determine the size of the Federal support mechanism. See also Texas PUC at 9 (support mechanism based on jurisdictional separations process must be converted to a more explicit system); NCTA at 7 (Part 36 separation procedures to support a USF and use of DEM weighting should be phased out as quickly as possible); Illinois CC at 9 (encourages FCC to move away from separations mechanisms as a means to achieve high cost funding).

[48] A component of existing ILEC prices that must also be recovered on an explicit basis is the cost caused by past regulatory intervention in depreciation practices. ILECs must be permitted to recover the costs of their embedded networks that were constructed in good faith with the expectation of eventual recovery. This program should be separate from any ongoing support for COLRs.

[49] The South Dakota Public Utilities Commission says at 2: "Expansion of the [EUCL], or alternatively, full rate recognition of customer-assignable and customer-specific non-traffic sensitive costs will increase explicit subsidies necessary to assure universal service, comparable service, and comparable rates. This truth is more critical in a rural and relatively low-income state such as South Dakota."

[50] Specific reductions should be proposed by the ILEC and accepted by each state. Rate design issues will vary widely across the states.

[51] See also NCTA at 13, that suggests that state toll revenues, as well as access, should be considered.

[52] See USTA at Attachment One.

[53] GTE proposed that the "laboratory model," as identified in the KickStart Initiative, Connecting America's Communities to the Information Superhighway, be used as the starting point. This model would provide every qualifying educational entity with connectivity to the Information Superhighway, yet not be overly burdensome on contributors. See GTE's comments at n.33 and Appendix D. The laboratory model would provide an equal initial level of connectivity. After completion of this initiative, further functionality could be added as found necessary and as funding permitted. See also USTA at 8-10.