*Pages 1--67 from Microsoft Word - 22280.doc* Federal Communications Commission FCC 02J- 2 Before the Federal Communications Commission Washington, D. C. 20554 In the Matter of Federal- State Joint Board on Universal Service ) ) ) ) CC Docket No. 96- 45 RECOMMENDED DECISION Adopted: October 15, 2002 Released: October 16, 2002 By the Federal- State Joint Board on Universal Service: Commissioners Abernathy, Copps and Consumer Advocate Gregg issuing separate statements; Commissioners Martin and Dunleavy approving in part, dissenting in part, and issuing separate statements; Commissioner Rowe dissenting and issuing a statement. TABLE OF CONTENTS Paragraph RECOMMENDED DECISION ................................................................................................................................. 1 I. INTRODUCTION.............................................................................................................................................. 2 II. BACKGROUND ................................................................................................................................................ 2 III. DISCUSSION..................................................................................................................................................... 7 A. SUFFICIENCY.................................................................................................................................................... 7 A. USE OF COSTS RATHER THAN RATES TO DETERMINE NON- RURAL HIGH- COST SUPPORT............................... 9 B. USE OF STATEWIDE AVERAGING TO REFLECT APPROPRIATE FEDERAL AND STATE ROLES IN ACHIEVING RATE COMPARABILITY ............................................................................................................................................ 11 C. BENCHMARK.................................................................................................................................................. 13 D. REASONABLE COMPARABILITY AND STATE INDUCEMENTS ........................................................................... 20 IV. RECOMMENDING CLAUSE ....................................................................................................................... 28 A. COST- BASED SUPPORT..................................................................................................................................... 1 B. RATES- BASED SUPPORT................................................................................................................................. 14 C. TRANSPORT ISSUE.......................................................................................................................................... 21 APPENDIX A (State Cost Cluster Analysis) APPENDIX B (Parties Filing Initial Comments) APPENDIX C (Parties Filing Reply Comments) 1 Federal Communications Commission FCC 02J- 2 2 I. INTRODUCTION 1. In this Recommended Decision, the Federal- State Joint Board on Universal Service (Joint Board) provides its recommendations on issues from the Ninth Report and Order that were remanded to the Commission by the United States Court of Appeals for the Tenth Circuit. 1 The Ninth Report and Order established a federal high- cost universal service support mechanism for non- rural carriers based on forward- looking economic costs. Consistent with the court’s decision, the Joint Board recommends that the Commission modify the non- rural high- cost support mechanism implemented in the Ninth Report and Order by adopting additional measures to induce states to ensure reasonable comparability of urban and rural rates. We also recommend that the Commission implement a supplementary rate review as a check on whether non- rural high- cost support continues to provide sufficient support to enable the states to maintain reasonably comparable rural and urban rates. In addition, we recommend continued use of statewide average costs to determine non- rural high- cost support. We believe that these recommendations will enable the Commission to satisfy the court’s remand and continue to fulfill Congress’s directive in the Telecommunications Act of 1996 to preserve and advance universal service. 2 II. BACKGROUND 2. The 1996 Act codified the commitment of the Commission and the state regulators to promote universal service in order to help ensure that consumers in all regions of the nation have access to affordable, quality telecommunications services. 3 In section 254 of the Act, Congress directed the Commission, after consultation with the Joint Board, to establish specific, predictable, and sufficient support mechanisms to preserve and advance universal service. 4 In addition, in section 254( b), Congress provided a list of principles upon which the Commission must base policies for the preservation and advancement of universal service. 5 Among other things, section 254( b) states that consumers in rural, insular, and high- cost areas should have 1 Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Ninth Report and Order and Eighteenth Order on Reconsideration, 14 FCC Rcd 20432 (1999) (Ninth Report and Order) remanded, Qwest Corp. v. FCC, 258 F. 3d 1191 (10th Cir. 2001). 2 Telecommunications Act of 1996, Pub. L. No. 104- 104, 110 Stat. 56 (1996) (1996 Act). The 1996 Act amended the Communications Act of 1934. 47 U. S. C. §§ 151, et seq. (Communications Act or Act). References to section 254 in this Recommended Decision refer to the universal service provisions of the 1996 Act, which are codified at 47 U. S. C. § 254 of the Act. See 47 U. S. C. § 254( b). 3 47 U. S. C. § 254( b). 4 Id. at §§ 254( a), (b)( 5), (d), (e). See also Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Notice of Proposed Rulemaking and Order Establishing Joint Board, 11 FCC Rcd 18092 (1996) (Universal Service NPRM). 5 47 U. S. C. § 254( b). 2 Federal Communications Commission FCC 02J- 2 3 access to telecommunications services at rates that are “reasonably comparable to rates charged for similar services in urban areas.” 6 3. Based on recommendations from the Joint Board 7 and building on the framework set forth by the Commission in prior orders, 8 the Commission adopted the Ninth Report and Order on October 21, 1999, establishing a federal high- cost universal service support mechanism for non- rural carriers based on forward- looking economic costs. 9 With the Ninth Report and Order, the Commission sought to “adopt a new specific and predictable forward- looking mechanism that will provide sufficient support to enable affordable, reasonably comparable intrastate rates for customers served by non- rural carriers.” 10 6 Id. at § 254( b)( 3). 7 See, e. g., Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Second Recommended Decision, 13 FCC Rcd 24744 (Jt. Bd. 1998) (Second Recommended Decision). 8 See, e. g., Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Report and Order, 12 FCC Rcd 8776 (1997), as corrected by Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Erratum, FCC 97- 157 (rel. June 4, 1997), aff’d in part, rev’d in part, remanded in part sub nom. Texas Office of Public Utility Counsel v. FCC, 183 F. 3d 393 (5 th Cir. 1999), petition for stay granted in part (Sept. 28, 1999), petitions for rehearing and rehearing en banc denied (Sept. 28, 1999) (First Report and Order); Federal- State Joint Board on Universal Service, Access Charge Reform, Seventh Report & Order and Thirteenth Order on Reconsideration in CC Docket No. 96- 45, Fourth Report & Order in CC Docket No. 96- 262, and Further Notice of Proposed Rulemaking, 14 FCC Rcd 8077 (1999), petition for review filed sub nom. Vermont Department of Public Service v. FCC, No. 99- 60530 (5 th Cir., filed June 23, 1999) (Seventh Report and Order). 9 Ninth Report and Order, 14 FCC Rcd at 20439, para. 2. Non- rural carriers are those that do not meet the following statutory definition of a rural telephone company: The term “rural telephone company” means a local exchange carrier operating entity to the extent that such entity-- (A) provides common carrier service to any local exchange carrier study area that does not include either- (i) any incorporated place of 10,000 inhabitants or more, or any part thereof, based on the most recently available population statistics of the Bureau of the Census; or (ii) any territory, incorporated or unincorporated, included in an urbanized area, as defined by the Bureau of the Census as of August 10, 1993; (B) provides telephone exchange service, including exchange access, to fewer than 50,000 access lines; (C) provides telephone exchange service to any local exchange carrier study area with fewer than 100,000 access lines; or (D) has less than 15% of its access lines in communities of more than 50,000 on February 8, 1996. 47 U. S. C. § 153( 37). See also First Report and Order, 12 FCC Rcd at 8944, para. 310. 10 Ninth Report and Order, 14 FCC Rcd at 20451, para. 34. 3 Federal Communications Commission FCC 02J- 2 4 4. The forward- looking mechanism implemented in the Ninth Report and Order determines the amount of federal support to be provided to non- rural carriers in each state by comparing the statewide average cost per line for non- rural carriers, as estimated by the Commission’s cost model, to a nationwide cost benchmark. 11 The Commission adopted the Joint Board’s recommendation to use costs as a proxy for rates in assessing its responsibilities to enable the reasonable comparability of rates. 12 The Commission concluded that comparing costs in different states, rather than rates, would allow the federal mechanism to provide sufficient support to enable reasonably comparable rates without having to evaluate the myriad state policy choices that affect those rates. 13 The Commission stated that “if federal support is available to cover costs that substantially exceed the national average, no state should face rates that are significantly higher than those elsewhere.” 14 5. The Commission determined that, for purposes of determining non- rural support, the statewide averaging approach was most consistent with the federal role of providing support for intrastate universal service to enable the states to ensure reasonable comparability of rates. 15 The Commission acknowledged that states set intrastate rates and, therefore, hold the responsibility of ensuring reasonable comparability of rates within their borders. 16 The federal mechanism operates by transferring funds among jurisdictions and has the effect of shifting money from relatively low- cost states to relatively high- cost states. 17 No state with forward- looking costs greater than the national benchmark would be forced to keep rates reasonably comparable without the benefit of federal support. 18 6. The non- rural mechanism provides support for the percentage of the costs per line allocated to the intrastate jurisdiction that exceed a national average cost benchmark of 135%. 19 11 The cost model estimates the forward- looking costs of providing supported services for non- rural carriers. The Commission selected input values for the model in the Tenth Report and Order, and found the model provides reasonably accurate cost estimates. Federal- State Joint Board on Universal Service, Forward- Looking Mechanism for High Cost Support for Non- Rural LECs, CC Docket Nos. 96- 45, 97- 160, Tenth Report and Order, 14 FCC Rcd 20156 (1999) (Tenth Report and Order), affirmed, Qwest Corp. v. FCC, 258 F. 3d 1191 (10th Cir. 2001). 12 Ninth Report and Order, 14 FCC Rcd at 20441, para. 15. 13 Id. at 20447, para. 25. 14 Id. 15 Id. at 20457, para. 45. 16 Id. at 20458, para. 46. 17 Id. at 20457, para. 45. 18 Id. 19 See id. at 20467, para. 63. The federal high- cost universal service support mechanism for non- rural carriers provides support for 76% of the costs that are above the national benchmark. The forward- looking mechanism calculates support based on 75% of forward- looking loop costs and 85% of forward- looking port costs, as well as 100% of all other forward- looking costs determined by the Commission’s forward- looking high- cost model. Based (continued….) 4 Federal Communications Commission FCC 02J- 2 5 The Commission concluded in the Ninth Report and Order that a benchmark of 135% of the national average cost balanced various goals under the statute, including sufficiency and the need to achieve rate comparability. 20 In addition, the Commission attempted to ensure that the fund would be no larger than necessary in order to minimize burdens on carriers and consumers contributing to universal service mechanisms. 21 7. In its remand of the Ninth Report and Order, the court determined that the Commission did not adequately explain its decision in certain respects. 22 The court observed that the Commission must base its universal service policies on the principles listed in section 254( b). In particular, the court found two principles in section 254( b) most relevant to the case: the principle that consumers in “rural, insular, and high cost areas” should have access to services “that are available at rates that are reasonably comparable to rates charged for similar services in urban areas;” 23 and the principle that “[ t] here should be specific, predictable and sufficient Federal and State mechanisms to preserve and advance universal service.” 24 The court also noted section 254( e), which states that any federal support for universal service “should be explicit and sufficient to achieve the purposes of this section.” 25 8. The court determined that the Commission did not provide an adequate explanation for its decision that the federal high- cost universal service support mechanism for non- rural carriers achieved the statutory principles codified in section 254 of the Act. 26 Without such an explanation, the court stated that it could not review the rationality of the Ninth Report and Order. 27 The court explained: “We do not decide the underlying issue of whether the funding is in fact sufficient; rather we conclude that the FCC has not supported why the funding is sufficient.” 28 The court remanded the Ninth Report and Order to the Commission to “establish an adequate legal and factual basis for the Ninth Order and, if necessary, to reconsider the operative mechanism promulgated in that Order.” 29 In particular, the court concluded that the (Continued from previous page) on the percentage of forward- looking costs that the intrastate portion of each of the items represents, the Commission determined that together the items represent 76% of total forward- looking costs. Id. 20 Id. at 20464, para 55. 21 Id. 22 Qwest Corp. v. FCC, 258 F. 3d at 1195. 23 47 U. S. C. § 254( b)( 3). 24 Id. at § 254( b)( 5). 25 Id. at § 254( e). 26 Qwest, 258 F. 3d at 1202. 27 Id. at 1205. 28 Id. at 1195. 29 Id. 5 Federal Communications Commission FCC 02J- 2 6 Commission did not: (1) define adequately the key statutory terms “reasonably comparable” and “sufficient”; (2) adequately explain setting the funding benchmark at 135% of the national average; (3) provide inducements for state universal service mechanisms; or (4) explain how this funding mechanism will interact with other universal service programs. 30 9. On February 15, 2002, the Commission issued the Remand Notice seeking comment on the first three issues remanded by the Tenth Circuit and referring the record collected in the proceeding to the Joint Board for a recommended decision. 31 In the Remand Notice, the Commission reserved review of the fourth issue on remand. 32 The Commission stated that a response to the remanded issues relating directly to the non- rural mechanism and a critical examination of the mechanism should be completed prior to a comprehensive review of the rural and non- rural universal service support mechanisms. 33 As a result, the Joint Board recommendations outlined in this decision apply to the non- rural high- cost universal service support mechanism and do not address the rural mechanism. 34 As the Commission seeks to better coordinate and reconcile the rural and non- rural mechanisms, it may find it necessary to review or adjust the principles and procedures developed herein in response to the Tenth Circuit remand. 30 Id. at 1201. 31 See Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Notice of Proposed Rulemaking and Order, 17 FCC Rcd 2999, 3010- 11, paras. 25- 26 (2002) (Remand Notice). 32 See id. at 3010, para. 25. 33 See id. 34 Currently there are two separate mechanisms that provide high- cost universal service support: the forward-looking economic cost mechanism for non- rural carriers; and the modified embedded cost mechanism for rural carriers. When the Commission determined in May 1997 that universal service support should be based on forward- looking economic cost, it decided to implement such a mechanism first for non- rural carriers. See First Report and Order, 12 FCC Rcd at 8889, para. 203. Because rural carriers generally have higher operating and equipment costs, which are attributable to lower subscriber density, small exchanges, and a lack of economies of scale, the Commission recognized that additional effort would be needed to develop a forward- looking mechanism appropriate for rural carriers. To assist in this challenge, the Joint Board established the Rural Task Force, which was comprised of individuals representing rural telephone companies, competitive local exchange carriers, interexchange carriers, wireless providers, consumer advocates, and state and federal agencies. Rather than attempting to modify the Commission’s forward- looking cost mechanism that currently is used to determine non-rural support, the Rural Task Force proposed modifications to the current embedded cost system for a five- year period. See Letter from William R. Gillis, Chair, Rural Task Force, to Magalie Roman Salas, FCC, dated September 29, 2000. In May 2001, consistent with the Joint Board’s recommendation, the Commission modified its rules for providing high- cost universal service support to rural telephone companies for five years based upon the proposals made by the Rural Task Force. See Federal- State Joint Board on Universal Service, Multi- Association Group (MAG) Plan for Regulation of Interstate Services of Non- Price Cap Incumbent Local Exchange Carriers and Interexchange Carriers, Fourteenth Report and Order, Twenty- Second Order on Reconsideration, and Further Notice of Proposed Rulemaking in CC Docket No. 96- 45, Report and Order in CC Docket No. 00- 256, 16 FCC Rcd 11244, 11257, para. 27 (2001) (Rural Task Force Order). 6 Federal Communications Commission FCC 02J- 2 7 III. DISCUSSION 10. Based on examination of the record in this proceeding, the Joint Board recommends that the Commission modify the non- rural high- cost support mechanism implemented in the Ninth Report and Order by adopting additional measures that will establish specific inducements for states to ensure that rates in all regions of the nation are reasonably comparable to rates in urban areas. We also recommend that the Commission implement a supplementary rate review to assess whether non- rural high- cost support continues to provide sufficient support to enable the states to maintain reasonably comparable rates. Consistent with the court’s decision, our recommendations with regard to these additional measures will support and complement the Commission’s initial decision in the Ninth Report and Order. Specifically, we recommend a process that includes the following: (1) continuing use of a national average cost benchmark based on 135% of the national average cost; (2) funding 76% of state average costs exceeding the national benchmark; (3) establishing a national rate benchmark based on a percentage of the national average urban rate; (4) implementing state review and certification of rate comparability; and (5) providing states the opportunity to demonstrate that further federal action is needed because current federal support and state actions together are insufficient to yield reasonably comparable rates. 11. The Joint Board’s recommendations comprise an integrated approach to the complex and interrelated issues referred by the Commission. We believe that these recommendations will enable the Commission to satisfy the court’s remand and continue to fulfill Congress’s directive to preserve and advance universal service. We note that this mechanism calculates support only for non- rural carriers. Certain assumptions in this Recommended Decision may not make sense for rural carriers. For example, as discussed below, while statewide averaging is appropriate in the non- rural mechanism, it may not be appropriate for the high- cost mechanism providing support to rural carriers. 35 A. Sufficiency 1. Background 12. Section 254( b)( 5) of the Act provides that “[ t] here should be specific, predictable and sufficient Federal and State mechanisms to preserve and advance universal service.” 36 Section 254( e) provides that federal support “should be explicit and sufficient to achieve the purposes of section 254[.]” 37 In the Ninth Report and Order, the Commission stated that the non- rural high-cost support mechanism would “provide sufficient support to enable reasonably comparable rates.” 38 The Commission also stated that the level of support would be “sufficient to ‘prevent 35 See infra para. 28. 36 47 U. S. C. § 254( b)( 5). 37 Id. at § 254( e). 38 Ninth Report and Order, 14 FCC Rcd at 20464, para. 56. 7 Federal Communications Commission FCC 02J- 2 8 pressure from high costs and the development of competition from causing unreasonable increases in rates above current, affordable levels. ’” 39 13. The court found that the Commission did not define the key statutory term “sufficient,” but simply asserted without explanation that the mechanism it chose would be sufficient. 40 The court declared the rationale conclusory and, thus, “inadequate to enable appellate review of the sufficiency of the federal mechanism.” 41 The court required the Commission on remand to define “sufficient” more precisely “in a way that can be reasonably related to the statutory principles, and then to assess whether the funding mechanism will be sufficient for the principle of making rural and urban rates reasonably comparable.” 42 14. The relationship between “sufficiency” and fund size has been addressed by the Joint Board and the Commission in other decisions. In the Second Recommended Decision, the Joint Board recommended a non- rural high- cost support mechanism that would enable rates to remain affordable and reasonably comparable, “but that is no larger than necessary to satisfy that statutory mandate.” 43 The Joint Board stated that the correct fund size would be essential to ensuring that all consumers benefit from universal service. Thus, the Joint Board recognized that, in implementing a non- rural high- cost universal service mechanism, it must be mindful of two competing goals: “( 1) supporting high- cost areas so that consumers there have affordable and reasonably comparable rates; and (2) maintaining a support system that does not, by its sheer size, overburden consumers across the nation.” 44 The Commission has indicated that “sufficiency” requires that universal service support not be excessive, citing the United States Court of Appeals for the Fifth Circuit’s caution that “excessive funding may itself violate the sufficiency requirements of the Act.” 45 2. Discussion 15. The Joint Board recommends that, for purposes of non- rural high- cost support, sufficiency should be principally defined as enough support to enable states to achieve reasonable comparability of rates. Sufficiency should be defined based on the relevant statutory goals under section 254( b). Thus, the definition of the term may vary depending on the 39 Id. at 20446, para. 24 (quoting Seventh Report and Order, 14 FCC Rcd at 8092, para. 30). 40 Qwest Corp. v. FCC, 258 F. 3d 1191, 1201( 10 th Cir. 2001). 41 Id. at 1201. 42 Id. at 1202. 43 Second Recommended Decision, 13 FCC Rcd at 24746, para 3. 44 Id. 45 Rural Task Force Order, 16 FCC Rcd at 11257, para. 27 (quoting Alenco Communications, Inc. v. FCC, 201 F. 3d 608, 619 (5 th Cir. 2000)). 8 Federal Communications Commission FCC 02J- 2 9 underlying purpose of the universal service program in question. The principal purpose of the non- rural high- cost support mechanism is to provide enough federal support to enable states to achieve reasonable comparability of rural and urban rates, the principle found in section 254( b)( 3). As discussed in more detail below, non- rural high- cost support is designed to provide high- cost states enough support so that their net average costs are reasonably comparable to the national average cost. With reasonably comparable net costs, these high- cost states should then have the resources to ensure that rural and urban rates within their borders are reasonably comparable. The Joint Board recommends below that the Commission require states to certify that their rates are reasonably comparable or explain why they are not, and provide states the opportunity to demonstrate that further federal action is needed because current federal support and state action together are insufficient to achieve reasonably comparable rates. Accordingly, for purposes of non- rural high- cost support, the Joint Board recommends that sufficiency be defined as enough support to enable states to achieve reasonably comparable rates. 16. The Joint Board also reaffirms that the statutory principle of sufficiency means that non- rural high- cost support should be only as large as necessary to achieve its statutory goal. Correct fund size is essential to ensure that all consumers benefit from universal service. A. Use of Costs Rather Than Rates to Determine Non- Rural High- Cost Support 1. Background 17. In the Second Recommended Decision, the Joint Board recommended that, because rate setting methods and goals may vary across jurisdictions, the Commission “use the cost of providing all supported services, rather than local rates” to determine federal high- cost support. 46 The Joint Board explained that using costs as an indicator of a state’s ability to maintain reasonable comparability of rates was appropriate because states have broad discretion to develop local rate designs. 47 State rate designs may reflect a broad array of policy choices that affect local rates, including implicit intrastate subsidies, enhanced service and other intrastate services. 48 In the Seventh Report and Order, the Commission adopted the Joint Board’s recommendation to use costs as a proxy for rates to perform its responsibility to enable the reasonable comparability of rates. 49 2. Discussion 18. We explain more fully here why costs rather than rates should continue to be the principal basis for determining federal support flows among states. Congress adopted section 46 Second Recommended Decision, 13 FCC Rcd at 24754, para. 19. 47 Id. See 47 U. S. C. § 152( b). 48 Seventh Report and Order, 14 FCC Rcd at 8092- 3, para 32. 49 See id. at 8092- 3, paras. 32- 33. 9 Federal Communications Commission FCC 02J- 2 10 254 to ensure that, as competition develops, there would be explicit support mechanisms in place to preserve the fundamental communications policy goal of providing universal telephone service in all regions of the nation at reasonably comparable rates. 50 Section 254( b)( 3) requires reasonably comparable rates. 51 This would be a relatively easy undertaking if the cost of providing telephone service were comparable in urban and rural areas. But costs are not comparable. The cost of providing telephone service is largely a function of population density and distance. Sparsely populated, rural areas have longer telephone loops, the most expensive portion of the telephone network, and fewer customers to spread the costs among. In some rural areas the cost of providing telephone service may be one hundred times greater than costs in urban areas. 52 19. Although rates generally are related to costs, states may base rates on numerous considerations in addition to cost. For example, local rates may vary from state to state depending upon each state’s local rate design policies; whether or not a carrier’s rates are set based on a price cap approach; the degree to which implicit subsidies may remain within local rates; whether a state universal service fund exists; and other factors. Attempting to develop cost support levels based principally on rates would therefore likely be difficult to implement considering the lack of uniformity in local rate design practices and could lead to inequitable treatment between states with substantially similar costs but different local rate policies. 20. For these reasons, the use of costs rather than rates to determine federal support was central to the Commission’s decision adopting the non- rural high- cost support mechanism in the Ninth Report and Order. 53 We agree with the Commission’s past decision that cost analysis offers advantages over rate analysis for purposes of determining federal support levels. Cost analysis enables accurate comparison of states for purposes of determining federal support levels. The Commission has stated that “[ a] state facing costs substantially in excess of the 50 See S. REP. NO. 23, 104 th Cong., 1 st Sess. 25 (1995). 51 47 U. S. C. § 254( b)( 3). 52 See Federal- State Joint Board on Universal Service, Multi- Association Group (MAG) Plan for Regulation of Interstate Services of Non- Price Cap Incumbent Local Exchange Carriers and Interexchange Carriers, CC Docket No. 00- 256, Second Report and Order and Further Notice of Proposed Rulemaking, Federal- State Joint Board on Universal Service, CC Docket 96- 45, Fifteenth Report and Order, Access Charge Reform for Incumbent Local Exchange Carriers Subject to Rate- of- Return Regulation, CC Docket 98- 77, Report and Order, Prescribing the Authorized Rate of Return from Interstate Services of Local Exchange Carriers, CC Docket 98- 166, Report and Order, 16 FCC Rcd 19613, 19635- 6, para. 45 (2001)( MAG Order), recon. pending. This is based on comparing the forward- looking cost estimates of wire centers in the Commission’s cost model. In a case where the cost in a rural area is one hundred times greater than costs in urban areas, rural and urban rates that are within 70 to 80% may not seem unreasonable. For example, if the urban cost is $20, the corresponding rural cost that is one hundred times greater would be approximately $2000. An urban rate that reflects the urban cost would be approximately $20 per line, per month. The rural rate within 70 to 80% of the urban rate is $34-$ 36, rather than $2000. 53 See Ninth Report and Order, 14 FCC Rcd at 20453- 4, paras. 36- 38. See also, e. g., Verizon telephone companies (Verizon) Comments at 8. 10 Federal Communications Commission FCC 02J- 2 11 national average may be unable through any reasonable combination of local rate design policy choices to achieve rates reasonably comparable to those that prevail nationwide.” 54 Examining the underlying costs enables the Commission to “evaluate the cost levels that must be supported in each state in order to develop reasonably comparable rates.” 55 21. While the inducements to state action on rates and supplemental rate review contained in this recommendation recognize that the ultimate test of rate comparability will be the rates customers actually pay for service, the use of costs for determining the areas of greatest need establishes a firm foundation for the states to fulfill the goals of section 254 of the Act. We recommend that the Commission continue to use a cost- based approach as the principal means of achieving the statutory goal of rate comparability. B. Use of Statewide Averaging to Reflect Appropriate Federal and State Roles in Achieving Rate Comparability 1. Background 22. The non- rural high- cost support mechanism calculates support levels for non- rural carriers by comparing the forward- looking costs of providing supported services, averaged over the statewide level, to the national benchmark. In the Ninth Report and Order, after consultation with the Joint Board, the Commission concluded that statewide averaging would be most consistent with the federal role of providing support for intrastate universal service to enable reasonable comparability of rates among states. 56 By averaging costs at the statewide level, the federal mechanism compares the relative costs of providing supported services in different states. It then provides support to carriers in those states with costs that exceed 135% of the national average. As the Commission explained in the Ninth Report and Order, “[ t] his has the effect of shifting money from relatively low- cost states to relatively high- cost states . . . [and] ensures that no state with costs greater than the national benchmark will be forced to keep rates reasonably comparable without the benefit of federal support.” 57 Statewide averaging assigns to the states the primary responsibility for ensuring reasonably comparable rural rates within their borders and permits states to use their resources to achieve the goal of reasonable comparability within states. 58 54 Seventh Report and Order, 14 FCC Rcd at 8092- 3, para. 32. 55 Id. at 8092, para. 30. The Commission has previously noted that current rates produced under the existing support regime are affordable and reasonably comparable. See id. 56 Ninth Report and Order, 14 FCC Rcd at 20457, para. 45. 57 Id. 58 Id. at 20458, para 46. 11 Federal Communications Commission FCC 02J- 2 12 23. The court stated that it did not object to the Commission’s comparison of statewide and national averages to achieve the goals of the Act. 59 The court also recognized that the 1996 Act “plainly contemplates a partnership between the federal and state governments to support universal service.” 60 2. Discussion 24. The Joint Board recommends that the Commission continue to determine high- cost support for non- rural companies by using statewide average costs. We believe that this reflects an appropriate division of federal and state responsibility for achieving rate comparability for non- rural companies. Because the states, not the Commission, set intrastate rates, the states have primary responsibility for ensuring reasonably comparable rural and urban rates. States tend to rely on either implicit or explicit mechanisms to transfer support from low- cost lines to high- cost lines within a state. 25. Despite implicit or explicit state support mechanisms, the low- cost areas of some states cannot balance their high- cost areas. Although such states could, through their own efforts, achieve reasonably comparable rates within their own boundaries, those rates would still be high relative to the national average because of the states’ high average costs. The Commission’s primary role is to identify those states that do not have the resources within their borders to support all of their high- cost lines. The non- rural high- cost support mechanism achieves this through the comparison of statewide average cost to a national cost benchmark. The averaging process provides a logical means to assess the relative extent to which states can support their high- cost areas by using resources from low- cost areas. By shifting funds to states with average costs above the national benchmark, the Commission provides federal support that is intended to enable high- cost states to set rates that are reasonably comparable to all rates across the nation. 26. The Commission explained in the Ninth Report and Order that the non- rural high-cost support mechanism “has the effect of shifting money from relatively low- cost states to relatively high- cost states.” 61 The Commission believed that its non- rural support mechanism ensured that no state with costs greater than the national benchmark would be forced to keep rates reasonably comparable without the benefit of federal support. Statewide averaging assigns to the states the primary responsibility for ensuring reasonable comparability of rates within their borders and permits states to use their resources to achieve the goal of reasonable comparability within states. 62 We continue to support these policies. 59 Qwest Corp. v. FCC, 258 F. 3d 1191, 1202 n. 9 (10 th Cir. 2001). 60 Id. at 1203. 61 Ninth Report and Order, 14 FCC Rcd at 20457, para. 45. 62 Id. at 20458, para 46. 12 Federal Communications Commission FCC 02J- 2 13 27. We disagree with the contention of the Rural Utilities Service that high- cost customers are being hidden by statewide averaging. 63 The Rural Utilities Service was concerned about the circumstance in which some customers have high costs but the state average is not high enough to qualify for support. The use of statewide average costs reflects what we believe to be an appropriate policy decision that in such cases the state has the primary responsibility and demonstrated ability to ensure rate comparability. Federal support is needed when the state, because of its high average cost, cannot solve such a problem without imposing an undue burden on its own ratepayers. 28. While statewide averaging is appropriate in the non- rural mechanism, it may not be appropriate for the high- cost mechanism providing support to rural carriers. 64 Many rural carriers lack the economies of scale and scope of the generally larger non- rural carriers, as the Rural Task Force established in documenting differences that exist between rural and non- rural companies. 65 The Commission has stated that it intends to ask the Joint Board to conduct a comprehensive review of the high- cost support mechanisms for rural and non- rural carriers as a whole to ensure that both mechanisms function efficiently and in a coordinated fashion. Accordingly, the Joint Board does not address the complex issues surrounding high- cost support for rural telephone companies in this Recommended Decision. The Joint Board emphasizes that the current recommendation is not intended to apply to rural companies. Now that the Joint Board has concluded its recommended decision on the issues in the court’s remand, we look forward to a Commission referral of a comprehensive review of the rural and non- rural high- cost support mechanisms. C. Benchmark 1. Background 29. Before the Ninth Report and Order, both rural and non- rural carriers were eligible for support under the Commission’s high- cost loop support mechanism, which provided increasing amounts of explicit support based on the amount by which a carrier’s loop costs, as reflected in its books, exceeded the national average. 66 Beginning with loop costs between 115% and 160% of the national average, the high- cost loop support mechanism provided carriers with more than 63 See Rural Utilities Service ex parte Comments. 64 The Commission did not refer to the Joint Board the issue of how the non- rural high- cost support mechanism will interact with other universal service support programs, including high- cost support for rural carriers. See Remand Notice, 17 FCC Rcd at 3011, para. 26 n. 93. 65 See Letter from William R. Gillis, Rural Task Force, to Magalie Roman Salas, Federal Communications Commission, dated September 29, 2000. 66 Ninth Report and Order, 14 FCC Rcd at 20440, para 13. See 47 C. F. R. § 36.601, et. seq. In the Rural Task Force Order, the Commission determined that rural carriers would continue to receive support under the high- cost loop support mechanism, as modified pursuant to the recommendations of the Rural Task Force and the Joint Board, for a period of five years. See Rural Task Force Order, 16 FCC Rcd at 11246, para. 1. 13 Federal Communications Commission FCC 02J- 2 14 200,000 lines support for 10% of their costs, and gradually more support as those costs exceeded 160% of the national average. Carriers with fewer than 200,000 lines were eligible for greater levels of support. Because it provides gradually more support for costs that exceed certain thresholds, or steps, above the national average, the high- cost loop support mechanism is often referred to as a “step function benchmark.” 67 30. The current non- rural high- cost support methodology provides support for universal-service- related intrastate costs (as determined by the model) that exceed 135% of the national average cost per line for non- rural carriers. The Commission, following the Joint Board’s recommendations, developed and implemented this methodology in a series of decisions culminating with the Ninth Report and Order. In the First Report and Order, the Commission adopted the Joint Board recommendation to use forward- looking economic costs as the basis for determining support. 68 In the Seventh Report and Order, the Commission concluded that the new support mechanism should compare the forward- looking costs of providing supported services to a national, cost- based benchmark. 69 It also adopted the Joint Board’s recommendation that the new support mechanism “should not be significantly larger than the current explicit support mechanism.” 70 In the Ninth Report and Order, the Commission concluded that federal support should be provided for universal service related to all intrastate costs above the national benchmark, and that the benchmark level should be set at 135%. 71 31. The Commission stated several reasons for setting the benchmark level at 135%. In the Second Recommended Decision, the Joint Board recommended a national benchmark in the range between 115 and 150% of the national average cost per line. 72 The Commission reasoned that a benchmark of 135% “falls within the range recommended by the Joint Board, and ensures that no state will face costs greater than 135% above the national average cost per line.” 73 The Commission also reasoned that a 135% benchmark “is consistent with the precedent of the existing support mechanism and the comments we have received. The current mechanism begins providing support for costs between 115 and 160 percent of the national average cost per line, based on carriers’ books, and the vast majority of non- rural carriers receive all their current support for costs in this range. The new national benchmark of 135% is near the midpoint of this 67 Ninth Report and Order, 14 FCC Rcd at 20466, para. 60. 68 First Report and Order, 12 FCC Rcd at 8888, para. 199. 69 Seventh Report and Order, 14 FCC Rcd at 8107- 8, paras. 61- 62. 70 Id. at 8102, 8112, paras. 48 and 70. (“ Given that telephone service currently is largely affordable, and any significant increase in the size of federal support for local rates appears unnecessary, we conclude that we should limit the size of the federal mechanism, as recommended by the Joint Board.”) 71 Ninth Report and Order, 14 FCC Rcd at 20438, para. 10. 72 Second Recommended Decision, 13 FCC Rcd at 24761- 2, para. 43. 73 Ninth Report and Order, 14 FCC Rcd 20463- 4, para. 55. 14 Federal Communications Commission FCC 02J- 2 15 range[,]” and “a reasonable compromise of commenters’ proposals.” 74 The Commission further stated that “a national benchmark of 135% strikes a fair balance between the federal mechanism’s responsibility to enable reasonable comparability of rates among states and the burden placed on below- benchmark states (and ratepayers) whose contributions fund the federal support mechanism.” 75 32. The court found that the Commission failed to adequately explain how the benchmark of 135% would achieve the goals of the Act. 76 The court found the Commission’s justifications in the Ninth Report and Order insufficient, stating that “[ m] erely identifying some range and then picking a compromise figure is not rational decision- making.” 77 The court directed the Commission to address relevant data and provide adequate record support and reasoning on remand. 78 33. Although the court rejected the Commission’s justification for the benchmark, the court noted that, “if, however, the FCC’s 135% benchmark actually produced urban and rural rates that were reasonably comparable … we likely would uphold the mechanism.” 79 In addition, the court recognized that the Commission’s determination of a benchmark “will necessarily be somewhat arbitrary” and acknowledged that the Commission is entitled to deference when drawing a line in the case of a reasoned decision based on the record. 80 2. Discussion 34. Based on examination of the record, the Joint Board continues to support the 135% benchmark. As noted above, the court appeared to consider the ability to produce reasonably comparable urban and rural rates as a key factor in supporting an appropriate cost benchmark. 81 As the court observed, although non- rural high- cost support is distributed based on a comparison of national and statewide average costs, the benchmark must be ultimately based on attainment of the statutory principle of reasonable comparability of urban and rural rates. 82 We have noted 74 Id. 75 Id. at 20465, para. 58. 76 Qwest Corp. v. FCC, 258 F. 3d 1191, 1202 (10 th Cir. 2001). 77 Id. at 1202. The court did not address the Commission’s reasoning that its choice of 135% benchmark fairly balanced state and federal responsibilities under the Act. 78 Id. at 1203. 79 Id. at 1202. 80 Id. 81 See supra text accompanying note 79. 82 Qwest, 258 F. 3d at 1202. 15 Federal Communications Commission FCC 02J- 2 16 that the Joint Board and Commission have found in prior rulings that current rates are affordable and reasonably comparable. 83 These findings are supported by a recent General Accounting Office (GAO) report. 84 Based on data contained in the GAO Report, it appears that six years after passage of the Act the national averages of rural, suburban and urban rates for residential customers diverge by less than two percent. 85 We believe that the comparability of average rural and urban rates supports continued use of the 135% cost benchmark. In addition, the Joint Board finds that the current benchmark is empirically supported by a cluster analysis and a standard deviation analysis. Both of these methods indicate that the 135% benchmark targets support to states with substantially higher average costs than other states, consistent with the purpose of non- rural high- cost support. 35. Verizon argues that the 135% benchmark is consistent with Congressional intent that federal support be sufficient to maintain the range of rates existing at the time the 1996 Act was adopted. 86 We agree with Verizon that one of the goals of the 1996 Act was to ensure that rates remain reasonably comparable as competition develops. 87 Congress was concerned that competition would erode implicit support and adopted section 254 to preserve and advance universal service. 88 Verizon argues further that rates have not changed substantially since 1996, so the range of existing rates, as reflected in the GAO Report, should be used to determine what is reasonably comparable. Because 95% of rates fall within two standard deviations of the mean, Verizon argues that rural rates within two standard deviations of urban rates should be 83 See Seventh Report and Order, 14 FCC Rcd at 8092, para. 30. 84 United States General Accounting Office, Telecommunications: Federal and State Universal Service Programs and Challenges to Funding (GAO- 02- 187, Feb. 4, 2002) (GAO Report). We will address any issues concerning the General Accounting Office study later in this recommendation. 85 See Letter from W. Scott Randolph, Director – Regulatory Affairs for Verizon Communications, to Marlene H. Dortch, Federal Communications Commission, dated June 26, 2002 (Verizon June 26 ex parte). 86 See Letter from W. Scott Randolph, Director – Regulatory Affairs for Verizon Communications, to Marlene H. Dortch, Federal Communications Commission, dated August 16, 2002 (Verizon August 16 ex parte). 87 See Verizon August 16 ex parte. Verizon suggests that there is nothing in the statute to indicate that Congress believed there was an immediate problem with rate comparability because Congress did not direct the states to revise existing rate structures. Id. 88 47 U. S. C. § 254( b). Had Congress determined that rates were not reasonably comparable at the time the Act was passed, it would have directed the Commission to implement a mechanism to adjust rates immediately. Rather, Congress directed the Commission to preserve and advance universal service through the principles listed under section 254( b). While debating this issue in Congress, Senator Pressler noted that: “the need to preserve widely available and reasonably priced telephone service is one of the fundamental concerns addressed in The Telecommunications Competition and Deregulation Act of 1995.” 141 CONG. REC. S7886 (1995). Likewise, Senator Dorgan stated that Congress’s intent was to “make it clear that universal service must be maintained[,]” 141 CONG. REC. S7951 (1995), and Representative Bonilla stated that “[ i] t is essential that our rural residents continue to have equal and affordable phone service.” 141 CONG. REC. H8497 (1995). See also Verizon August 16 ex parte. 16 Federal Communications Commission FCC 02J- 2 17 considered reasonably comparable. 89 Verizon points out that an analysis of the Commission’s cost model shows that two standard deviations translates approximately to a 135% cost benchmark. Thus, Verizon argues that rural rates within two standard deviations of urban rates should be considered reasonably comparable and that the cost benchmark level of 135% is justified because it is nearly equivalent to two standard deviations. 90 As discussed below, we agree. 36. The current benchmark is supported by a standard deviation analysis. Standard deviation is a commonly used statistical analysis that measures dispersion of data points from the mean of those data points. In a normal distribution, data points within two standard deviations of the mean will comprise approximately 95% of all data points. In other words, use of two standard deviations will identify data points that are truly outliers within the sample studied. 91 Verizon points out that both the Commission and state commissions have adopted this statistical approach as a standard for determining parity or comparability. 92 As applied to the cost of non-rural lines, the measurement of two standard deviations from the national average cost results in approximately 132% of the national average cost. 93 Based on this information, the Joint Board concludes that the 135% benchmark is a reasonable dividing line separating high- cost states from the remainder of average and low- cost states. 89 See Verizon August 16 ex parte. 90 See id. at 2. 91 The cost data are not normally distributed, because there are more low- cost, urban lines than high- cost, rural lines. We are interested in providing support to states with more high- cost lines, so it is appropriate to use the two standard deviation measurement to identify outliers even though this measurement may identify more than expected in a normal distribution. The current non- rural high- cost support mechanism provides support to eight out of 52 jurisdictions (50 states, the District of Columbia and Puerto Rico), or 15% of the jurisdictions. 92 Verizon August 16 ex parte at Appendix A (describing the use of two standard deviations as a standard by the New York Public Service Commission in Case 97- C- 0139 – Proceeding on Motion of the Commission to Review Service Quality Standards for Telephone Companies, April 29, 2002, Appendix K, Statistical Methodoligies, page 1; the Pennsylvania Public Utility Commission in Joint Petition of NEXTLINK Pennsylvania, Inc., RCN Telecommunications Services of Pennsylvania, Inc., Hyperion Telecommunications, Inc., ATX Telecommunications, Focal Communications Corporation of Pennsylvania, Inc., CTSI, Inc., MCI Worldcom, e. Spire Communications, and AT& T Communications of Pennsylvania, Inc., for an Order Establishing a Formal Investigation of Performance Standards, Remedies, and Operations Support Systems Testing fro Bell Atlantic-Pennsylvania, Inc., Opinion and Order, P- 00991643, Order Entered December 31, 1999, page 143; and the Federal Communications Commission in PA PUC Metrics Order, Attachment A- 3, Calculation of Parity and Benchmark Performance, page 1). 93 Letter from W. Scott Randolph, Director – Regulatory Affairs for Verizon Communications, to Marlene H. Dortch, Federal Communications Commission, dated July 25, 2002 (Verizon July 25 ex parte). Based on the proxy model, Verizon calculated the national average at $23.35 and the standard deviation at $3.74. Adding the national average cost ($ 23.35) and two standard deviations ($ 3.74 x 2 = $7.48), the total is $30.83 ($ 23.35 + $7.48). $30.83 is approximately 132% of the national average cost of $23.35. Id. at 8. 17 Federal Communications Commission FCC 02J- 2 18 37. The Joint Board used a cluster analysis to determine that the states receiving non-rural high- cost support under the current 135% benchmark are states that have substantially higher average costs than other states. Cluster analysis is an analytical technique that organizes information around variables so that relatively homogeneous groups, or clusters, can be identified. The Joint Board used cluster analysis to identify groups of states that had similar cost characteristics, thereby warranting different treatment regarding universal service support. Specifically, states were sorted from lowest- to highest- cost based on statewide average cost per loop. 94 Clusters were identified in this ranking if the difference in average costs between states was greater than “cluster split differences” ranging from 2.5 to 0.5. 95 Under this analysis, Mississippi was the first to break out into a separate cluster, and the second was the District of Columbia. The first group of states to break out into a separate rural, high- cost cluster included Kentucky, Maine, Alabama, Vermont, Montana, West Virginia and Wyoming. The remaining states, ranging from New Jersey to Nebraska, formed a separate urban, low- cost cluster. When Mississippi and the District of Columbia, the respective high- and low- cost “outliers,” were combined into the two larger clusters, “cluster stability” was achieved for a wide range of numerical values from 2.5 to 0.85. 96 “Cluster stability” means that the same clusters are maintained even as the numerical values are varied, indicating a strong similarity among members of the cluster groups. Because cluster analysis identifies a high- cost, rural cluster of states that matches the group of states currently receiving support under the non- rural high- cost support mechanism, the Joint Board finds that the cluster analysis empirically supports the current 135% benchmark. 38. Because the standard deviation analysis and the cluster analysis both support 135% as a reasonable benchmark, the Joint Board recommends continued use of the 135% benchmark. The court recognized that the use of any benchmark may be somewhat arbitrary; however, choice of a specific, percentage- based benchmark (as opposed to a mathematically calculated benchmark based on two standard deviations which may result in a different percentage each year) provides certainty to the funding process that carriers and states desire. Accordingly, the Joint Board recommends continued use of a 135% benchmark The supplemental rate comparability review which we recommend will allow the Commission to assess how successfully the non- rural high- cost support ensures reasonable comparability of rates. 39. Some commenters suggest that, in light of the court’s decision, it would be more appropriate to use a benchmark based on average urban cost, rather than nationwide average cost. 97 The Joint Board recommends that the Commission continue to use a nationwide cost 94 Once sorted, it was then possible to identify cluster with small cost differences among the states within the cluster. Cluster analysis accomplishes this by comparing differences between and among groups, and choosing the number of groups. 95 The numerical values used to measure the cluster split differences were 2. 5, 2. 0, 1. 55, 1. 5, 1. 0, 0. 85, and 0. 5. 96 Attached as Appendix A is a spreadsheet illustrating this analysis. 97 See Maine Public Service Commission, Montana Public Service Commission and Vermont Public Service Commission (Rural State Commissions) Comments. 18 Federal Communications Commission FCC 02J- 2 19 benchmark. The national benchmark is intended to ensure that each state has a relatively equal ability to achieve reasonable comparability of urban and rural rates. We do not agree that an urban cost benchmark would better satisfy the statutory comparison of urban and rural rates. Like the current mechanism, the urban benchmark substitutes costs for rates. In addition, rather than comparing rural and urban costs, it compares statewide average costs to nationwide urban costs. 98 40. The urban benchmark proposal would require more funding or a higher benchmark level because urban average costs are lower than national average costs. For example, an urban benchmark of 165% would yield roughly the same support amounts as the current 135% national benchmark. An urban benchmark of less than 165% would require more federal support. The GAO Report suggests that more federal support is not necessary because urban and rural rates are similar. Proponents of the urban benchmark have not explained how additional funding produced by an urban benchmark would produce reasonably comparable rates, nor have they provided a rational justification for setting the benchmark at any particular level. 41. The urban benchmark proposal is premised in part on the argument that the current 135% national benchmark cannot enable rate comparability because it is equivalent to about 165% of urban average cost, near the 70- 80% range of variability that the court doubted was reasonably comparable. As explained above, however, rates do not necessarily equate to costs, so setting a 135% national benchmark (or 165% urban benchmark) does not mean intrastate rates will vary to the same degree. For the same reason, establishing cost support based on an urban benchmark will not ensure that urban and rural rates will be reasonably comparable. Because the urban benchmark proposal does not improve the operation of the high- cost support mechanism, nor address the rate comparability concerns of the court, the Joint Board recommends that the current national benchmark be retained, supplemented by rate review to ensure comparability of urban and rural rates. 42. As discussed above, a “step function” provides gradually more support for costs that exceed certain thresholds or “steps” above the national average. 99 BellSouth supports the 135% benchmark, but proposes an additional, lower benchmark to provide some support to carriers in states with average costs between 100 and 135% of the national average cost. BellSouth proposes a step function as a means of distributing support more widely among states and, thereby, inducing states to ensure reasonable comparability of urban and rural rates. As discussed above, the purpose of non- rural high- cost support is to provide sufficient support to enable high- cost states to develop reasonably comparable rates. Providing additional support 98 We also note that, while discussing the use of national and statewide averages for the statutory comparison of urban and rural rates, the court rejected the “argument that the use of statewide and national averages is necessarily inconsistent with § 254.” Qwest Corp. v. FCC, 258 F. 3d 1191, 1202 n. 9 (10 th Cir. 2001). 99 See supra para. 29. For example, whereas the non- rural high- cost support mechanism supports all intrastate costs above the 135% national benchmark, carriers with more than 200,000 lines were formerly eligible for support for 10% of their costs between 115 and 160% of the national average and gradually more support for costs exceeding 160%. See 47 C. F. R. § 36.601 et seq. 19 Federal Communications Commission FCC 02J- 2 20 merely to induce states to ensure rate comparability without determining that additional support is necessary may conflict with the principle that support should be only as large as necessary. Nevertheless, a step function could promote predictability by preventing a total loss of federal support if small cost changes cause a state’s average cost per line to fall below the dollar amount of the 135% benchmark in a given year. 100 We believe that use of a step function may have benefits and warrants further consideration; however, the Joint Board does not recommend that the Commission add a step function to the non- rural high- cost support mechanism at this time. In light of the need to respond expeditiously to the court’s remand, the Joint Board expects to address the issue of a step function in its comprehensive review of the rural and non- rural support mechanisms. D. Reasonable Comparability and State Inducements 1. Background 43. Section 254( b)( 3) states that “[ c] onsumers in all regions of the Nation, including low-income consumers and those in rural, insular, and high- cost areas, should have access to telecommunications and information services … at rates that are reasonably comparable to rates charged for similar services in urban areas.” 101 In the Seventh Report and Order, the Commission adopted the Joint Board’s interpretation of “reasonably comparable” as “a fair range of urban/ rural rates both within a state’s borders, and among states nationwide.” 102 Noting that “[ t] he Joint Board and the Commission have concluded that current rate levels are affordable[,]” the Commission further explained that “we interpret the goal of maintaining a ‘fair range’ of rates to mean that support levels must be sufficient to prevent pressure from high costs and the development of competition from causing unreasonable increases in rates above current, affordable levels.” 103 In the Ninth Report and Order, the Commission also stated “reasonably comparable must mean some reasonable level above the national average forward- looking cost per line, i. e., greater than 100% of the national average.” 104 100 We note that small changes in a state’s average cost per line could cause complete loss of federal support for certain states regardless of the level of the benchmark and regardless of whether a step function was adopted. The potential “on/ off” nature of support in the high- cost support mechanism is a function of the use of a benchmark, not of the particular level of the benchmark. To the extent that a state’s average cost per line is at or near the 135% benchmark, however, an additional lower benchmark would prevent the state from losing all federal support if its costs go below 135%. 101 47 U. S. C. § 254( b)( 3). 102 Seventh Report and Order, 14 FCC Rcd at 8092, para. 30. 103 Id. 104 Ninth Report and Order, 14 FCC Rcd at 20463, para. 54. 20 Federal Communications Commission FCC 02J- 2 21 44. The court found the Commission’s definition of “reasonably comparable” inadequate, and required a more precise definition “that reasonably relates to the statutory principles[.]” 105 The court stated that the “fair range” definition does not help answer the questions that arise about reasonable comparability. For example, Vermont and Montana assert that some rural rates will be 70- 80% higher than urban rates under the FCC’s funding mechanism. We fail to see how the FCC’s definition of ‘reasonably comparable’ illuminates this dispute. Does the FCC contend, for example, that a 70- 80% discrepancy is within a ‘fair range’ of rates? We doubt that the statutory principle of ‘reasonabl[ e] comparab[ ility] ’ can be stretched that far. 106 The court also stated that the Commission’s further definitions were little more precise than the first and, in any event, were not reasonable interpretations of the statutory language: “The Act calls for reasonable comparability between rural and urban rates: these definitions simply substitute different standards.” 107 45. The court also required the Commission to develop mechanisms to induce state action to ensure reasonable comparability of rural and urban rates. 108 In the Ninth Report and Order, the Commission adopted the Joint Board’s recommendation that it “abstain from requiring any state action as a condition for receiving federal high- cost service support (other than state certifications)[.]” 109 The Commission found it most appropriate for states to determine how non-rural high- cost support is used, “[ b] ecause the support . . . is intended to enable the reasonable comparability of intrastate rates, and states have primary jurisdiction over intrastate rates[.]” 110 As a regulatory safeguard, the Commission required states that wish to receive non- rural high-cost support to certify annually that all such support will be used in a manner consistent with section 254( e). 111 46. The court, noting that the Act “plainly contemplates a partnership between the federal and state governments to support universal service[,]” agreed that “it is appropriate— even 105 Qwest Corp. v. FCC, 258 F. 3d 1191, 1202 (10 th Cir. 2001). 106 Id. at 1201. 107 Id. 108 Id. at 1203- 04. 109 Ninth Report and Order, 14 FCC Rcd at 20469- 70, para. 67. 110 Id. at 20482- 3, para. 95; see id. at 20483, para. 96 (“ As long as the uses prescribed by the state are consistent with section 254( e), we believe that the states should have the flexibility to decide how carriers use support provided by the federal mechanism.”) 111 Id. at 20483- 4, para. 97; see 47 C. F. R. § 54.313( a) (state must certify support “will be used only for the provision, maintenance, and upgrading of facilities and services for which the support is intended.”) 21 Federal Communications Commission FCC 02J- 2 22 necessary— for the FCC to rely on state action in this area.” 112 The court found fault, however, with the Commission’s reliance on states to “act on their own to preserve and advance universal service[,]” without inducing states to provide for rate comparability as required by the statute. 113 The certification process did not meet the court’s concerns because it failed to address rate comparability. 114 On remand, the court held that the Commission must “create some inducement— a ‘carrot’ or a ‘stick, ’ for example, or simply a binding cooperative agreement with the states— for the states to assist in implementing the goals of universal service.” 115 47. As stated above, the Joint Board and Commission determined previously that current rates generally are affordable, 116 and the Commission interpreted the goal of maintaining a “fair range” of rates to mean sufficient support to prevent “unreasonable increases in rates above current, affordable levels.” 117 Some commenters contend that the GAO Report shows that rates remain affordable and reasonably comparable. 118 Among other things, the General Accounting Office (GAO) found that there was no statistical difference in residential local telephone rates between central city, suburban, and rural places. 119 Other commenters contend that the GAO’s findings are not meaningful because the rates compared are not adjusted to account for the varying factors that may be included in rates in different jurisdictions. 120 48. The GAO gathered data on basic local telephone rates from state commissions for sampled locations throughout all fifty states and the District of Columbia. 121 Within each state, 112 Qwest, 258 F. 3d at 1203. The court, therefore, rejected Qwest’s argument “that the Commission alone must support the full costs of universal service.” Id. 113 Id. at 1203- 04. 114 Id. at 1203 n. 10 (“ It requires only that the state certify that the carrier is eligible to receive the federal funds and that the funds are being used for universal service as intended.”) 115 Id. at 1204. 116 Seventh Report and Order, 14 FCC Rcd at 8092, para. 30; First Report and Order, 12 FCC Rcd at 8780- 1, para. 2; Second Recommended Decision, 13 FCC Rcd at 24746, para. 3; Federal- State Joint Board on Universal Service, Recommended Decision, 12 FCC Rcd 87, 154, para. 133 (Jt. Bd. 1996) (First Recommended Decision). 117 Seventh Report and Order, 14 FCC Rcd at 8092, para. 30. 118 See GAO Report; Verizon Comments at 4- 6; AT& T Reply Comments at 3- 4. 119 See GAO Report at 15; Verizon Comments at 4; AT& T Reply Comments at 3- 4. 120 See Rural State Commissions Reply Comments at 3- 4. 121 For residential and single- line business customers, the GAO asked for the unlimited service rate and the message or measured service rate with the lowest rate. The rates do not include the Federal Subscriber Line Charge; state and local surcharges for items such as state universal service funding, 911 service, and taxes; the federal excise tax; or long distance fees and associated universal service charges and other taxes. Where offered, the GAO used the tariff rate for unlimited service. Where unlimited service is not available, the GAO calculated a monthly fee for (continued….) 22 Federal Communications Commission FCC 02J- 2 23 the GAO randomly selected places from three broad categories associated with population density: central city, suburban (other places within a metropolitan statistical area (MSA)), and rural (outside MSA). For most states, the GAO chose three places in each of the categories (and four or five in larger states). The GAO Report provides a chart showing the average residential rate as approximately $15.00. 122 49. In addition to the GAO Report, rate data is available from the Commission’s own sources. The Commission’s Wireline Competition Bureau (Bureau) has conducted an annual survey of local telephone rates in 95 urban areas for the past 15 years. 123 The most recent survey indicates that the average urban rate paid by residential customers for flat- rate touch- tone calling is $21.84. 124 Unlike the GAO Report, the average urban rate in the Bureau’s study includes federal and state subscriber line charges, and other fees which consumers must pay each month in order to receive basic telephone service. 2. Discussion 50. The Joint Board recommends that the Commission implement a procedure that will induce states to achieve reasonably comparable rates and enable the Commission to take additional action, if necessary, to achieve comparable rates. Specifically, the Joint Board recommends the Commission expand the current annual certification process under Section 254( e) of the Act to require states to certify that the basic service rates in high- cost areas served by eligible telecommunications carriers (ETCs) within the state are reasonably comparable to a national rate benchmark. For purposes of this state certification process, the Joint Board (Continued from previous page) a “representative customer” using the message rate. The GAO assumed that a “representative customer” makes 100 5- minute calls per month. 122 GAO Report at 17. Verizon calculated the mean rates based on the GAO data as follows: central city, $14.79; suburb, $15.00; non- MSA, $14.76. See Verizon June 26 ex parte. 123 The cities surveyed are those that were included in the Bureau of Labor Statistics Consumer Price Index in the first year the survey was done. 124 See Reference Book of Rates, Price Indices, and Household Expenditures for Telephone Service (Industry Analysis and Technology Div. July 2002) (Reference Book). The most recent survey includes data as of October 15, 2001. All 95 cities in the sample had flat- rate service available in this survey. The average rate includes a monthly charge for flat- rate service ($ 14.05), federal and state subscriber line charges ($ 5.03), additional monthly charges for touch- tone service ($ 0.06), and taxes, 911 and other charges ($ 2.70). The average urban rate paid by customers for message or measured service is $15.16. Message or measured service was also available in 88 of the 95 cities surveyed. This average rate includes a monthly charge for measured/ message service ($ 7. 27), federal and state subscriber line charges ($ 5.03), additional monthly charges for touch- tone service ($ 0.06), and taxes, 911 and other charges ($ 2. 81). In prior surveys, flat- rate service was not available in all the sampled cities, and therefore the Bureau calculated a “representative rate” for each city in order to calculate a national average rate for the 95 cities. The representative rate was the flat- rate service charge in those areas where this type of service is available. If flat-rate service was unavailable, the rate for measured/ message service was used, along with the charges associated with placing 100 5- minute, same- zone, business- day calls. The average representative rate for residential local service has gone from $17.70 in 1986 to $21.84 in 2001. Id. 23 Federal Communications Commission FCC 02J- 2 24 recommends that high- cost areas be defined as all wire centers with a line density less than 540 lines per square mile. 125 As part of the certification process, all states should be required to compare basic service rates based on a standard template. The Commission should also establish a “safe harbor” whereby a state whose rates are at or below a certain rate benchmark may certify that their basic service rates in high- cost areas are reasonably comparable without the necessity of submitting rate information. However, states would have the option of submitting additional data to demonstrate that other factors affect the comparability of their rates. If a state’s rates are more than the rate benchmark, the state could request further federal action based on a showing that federal support and state actions together were not sufficient to yield reasonably comparable basic service rates statewide. Further federal actions could include, but are not limited to, additional targeted federal support, or actions to modify calling scopes or improve quality of service where state commissions have limited jurisdiction. A state requesting further federal action must show that it has already taken all actions reasonably possible and used all available state and federal resources to make basic service rates reasonably comparable, but that rates nevertheless fall above the benchmark. A state whose basic service rates exceed the rate benchmark and that requests further federal action should be required to submit rate data in support of its certification, based on a basic service rate template. The Joint Board recognizes that it may be appropriate to use 135% for the safe harbor rate benchmark, but recommends that the Commission further develop the record to establish the appropriate rate benchmark for the safe harbor. 51. The Joint Board believes that this expanded certification process meets the court requirement to induce state action to achieve rate comparability. With any support mechanism, the proof of success must be evaluated not only on whether the mechanism as a whole generally achieves rate comparability, but also upon the degree and nature of any exceptions. The court criticized the Commission for failing to adequately reconcile its conclusion that rates were generally comparable in light of instances where state rates were reportedly high. 126 Together 125 Wire centers with fewer than 540 lines per square mile are above the national average cost, and those with more than 540 lines per square mile are below the national average cost, based on current data. In addition, the average costs of wire centers with fewer than 540 lines per square mile vary greatly. In order to assist states in making their certifications, we recommend that identification of wire centers within each state with fewer than 540 lines per square mile be provided by USAC. The Joint Board suggests that the Commission specifically solicit comments on whether a different definition of rural and high- cost areas may be more appropriate for purposes of the state rate review and certification process. 126 Petitioners claimed that some rural rates would be 70- 80% higher than urban rates under the Commission’s non- rural support mechanism. This assertion was based on comparing cost data generated by the Commission’s cost model. By defining urban costs as the average cost in wire centers with 50,000 or more lines, or in the alternative, 100,000 or more lines, petitioners estimated that average rural costs are 70- 80% above average urban costs. As noted above, when costs are not averaged, the differences between rural and urban costs are much greater, with costs in some rural wire centers 100 times (10,000%) above the average cost. See supra note 52. Neither petitioners nor any other party in the proceeding before the Commission submitted rate data in support of suggested benchmarks to ensure reasonably comparable rates. Because the Commission was simultaneously considering the inputs to its cost model while it was considering other aspects of the non- rural support mechanism, it urged interested parties to formulate their comments using the most recent cost model outputs available at that time. See Ninth Report and Order, 14 FCC Rcd at 20462- 3, para. 53. (continued….) 24 Federal Communications Commission FCC 02J- 2 25 with federal non- rural high- cost support, the expanded certification process will ensure that rates “… in all regions of the Nation… are reasonably comparable…” as set forth in section 254( b)( 3). 127 The expanded certification process encourages states to scrutinize their rates using the basic service rate template, to determine whether they are reasonably comparable, and if not, to take actions to make them reasonably comparable. When state basic service rates are at or below the rate benchmark level, then there should be a presumption that rates in that state are reasonably comparable to national urban rates. This recommended approach affords the states maximum flexibility to determine basic service rates. The Commission should accord substantial deference to these state certifications. i. Rate Benchmark 52. As an initial matter, the Joint Board recommends that the Commission base the rate benchmark on the most recent average urban residential rate as shown in the Bureau’s Reference Book, as modified to reflect the most recent changes in subscriber line charges (SLC). 128 The average urban rate can be adjusted annually based on data from the Bureau’s annual rate survey. The Joint Board recognizes that it may be appropriate to use 135% for the safe harbor benchmark. 129 Use of a 135% rate benchmark is consistent with the national average cost benchmark of 135%. 130 The Joint Board believes that, since cost- based support is provided to ensure statewide average costs do not exceed 135% of the national average, most states should be able to maintain average rates below 135% of the national average urban rate. Based on the current national average urban rate, as adjusted, a 135% rate benchmark would be $30.16 per line per month. 131 The Joint Board recommends that the Commission further develop the record to establish the appropriate rate benchmark for the safe harbor. 53. The Joint Board emphasizes that any rate benchmark established is meant simply as a “safe harbor” for the purposes of determining rate comparability. The Joint Board does not suggest through this Recommended Decision that it is appropriate that any rates be increased to (Continued from previous page) 127 47 U. S. C. 254( b)( 3). 128 Cost Review Proceeding for Residential and Single- Line Business Subscriber Line Charge (SLC) Caps, Access Charge Reform, Price Cap performance Review for Local Exchange Carriers, Order, CC Docket Nos. 96- 262 and 94- 1 (rel. June 5, 2002) (approving increase in cap on subscriber line charges from $5.00 to $6.00 based on cost justification). The average monthly increase in the SLC’s charged by non- rural companies effective July 1, 2002, was 50¢ per line. As a result, the average urban residential rates used in setting the rate benchmark should be adjusted upward to $22.34. 129 See Ohio Consumers’ Counsel, Maryland Office of People’s Counsel, Maine Public Advocate Office, Texas Office of Public Utility Counsel and Pennsylvania Office of Consumer Advocate Reply Comments at 5. 130 The national average urban rate of $21.84 per line is very close to the national average cost per line of $21.92 produced by the Commission’s cost model. 131 National average urban rate, adjusted, of $22.34 X 1.35 = rate benchmark of $30.16. This rate benchmark includes various charges in addition to the tariffed rate as set forth above. 25 Federal Communications Commission FCC 02J- 2 26 that level. The Joint Board recognizes and supports the role of state commissions in setting rates within each state. The Joint Board recommends requiring that states review only residential rate information at this time. The Joint Board suggests that it may be appropriate to solicit comment as to whether only residential or residential and business rates eventually should be reviewed by the states. ii. Basic Service Rate Template 54. The Joint Board recommends that the Commission establish a basic service rate template for states to use to compare rates. We suggest that the basic service rate template should include the items contained in the annual rate survey by the Bureau. The Joint Board recommends that the template include the following factors: the rate for a line with access to the public switched network, federal subscriber line charge, state subscriber line charge (if any), federal universal fund charge, state universal fund charge (if any), local number portability charge, telecommunications relay service charge, 911 charges, 132 federal universal service credits (if any), state universal service credits (if any) and the federal excise tax. iii. Expanded Rate Certification Process 55. The expanded state certification process would augment the existing state certification under section 254( e) of the Act. The existing procedure requires states to certify that all ETCs that receive federal universal service funding are using the funds to achieve the goals of the Act. The new procedure would expand reporting requirements to include a discussion of rate comparability. In the expanded certification process, states typically would report in one of four ways: a. Rates within the state fall below the benchmark and are considered by the state to be reasonably comparable. No further showing should be required. b. Rates are not below the benchmark, but may nevertheless be considered reasonably comparable. A state could show that due to other factors -- for example, additional services included in the basic service rate or the method in which the state has targeted existing universal service support -- the rates above the benchmark actually should be presumed reasonably comparable. In the alternative, the state could report on actions it intends to take to achieve reasonable comparability. c. Rates are below the benchmark, but are not reasonably comparable. A state may show that even though actual rates are within the safe harbor, the price paid for service received results in rates and services that are not reasonably 132 In states where 911 fees are not established on a statewide basis, the state should use a statewide average 911 fee for purposes of the standard rate template. Use of a statewide average will maintain the proper role of federal support for state, rather than local rates, and will reduce the number of separate rates in states where 911 fees are set locally. 26 Federal Communications Commission FCC 02J- 2 27 comparable. In this case, a state could show that existing basic service is lacking in some way. For example, the state could show that the local calling area size is too small to be considered comparable service, and that toll or extended area service charges should be included to produce a reasonably comparable rate. In addition to explaining why rates within the safe harbor should not be considered reasonably comparable, the state must also show the actions it has taken or is going to take to remedy the discrepancy, prior to requesting additional federal actions to achieve reasonably comparable rates. d. Rates are above the benchmark and are not reasonably comparable. A state could request federal action based on a showing that current combined federal and state actions are insufficient to produce reasonably comparable rates. If the state asserts that existing federal support and state resources are not sufficient for the state to attain reasonably comparable rates, the state should be required to show that it has already taken all available steps to remedy the situation, but that rates remain above the benchmark. If the state can make this showing, the Commission would consider taking further action to meet the needs of the state in achieving reasonably comparable rates. 133 56. The Joint Board recommends that states certifying that their rates fall at or below the national rate benchmark and are reasonably comparable should not be required to submit any additional rate information. Any states requesting additional federal action should be afforded great flexibility in making their presentations, but should be required to fully explain the basis for their request. Factors that should be addressed by any such state would include, but not be limited to: rate analysis and a demonstration why the state contends that rates are not reasonably comparable; any other factors that should be considered in evaluating rates; and a demonstration that the state has taken all reasonably possible steps to develop maximum support from within the state. The requesting state should fully explain how it has used any federal support currently received to help achieve comparable rates and whether the state has implemented a state universal service fund to support rates in high- cost areas of that state. The Joint Board recommends the Commission develop exact procedures to be used in filing and processing requests for further federal actions. In particular, the Joint Board recommends that the Commission establish a time limit for consideration of such state requests, to ensure that requests will be processed and decided expeditiously. 133 Existing per line cost support is portable to any ETC serving a customer in a high- cost wire center in a state receiving non- rural support. See 47 C. F. R. § 54.307. The exact amount of this high- cost support is available for inspection on USAC’s website and is updated quarterly. 27 Federal Communications Commission FCC 02J- 2 28 IV. RECOMMENDING CLAUSE 57. For the reasons discussed herein, this Federal- State Joint Board pursuant to section 254( a)( 1) and section 410( c) of the Communications Act of 1934, as amended, 47 U. S. C. §§ 254( a)( 1) and 410( c), recommends that the Commission adopt the proposals describe above relating to issues from the Ninth Report and Order that were remanded to the Commission by the United States Court of Appeals for the Tenth Circuit. FEDERAL COMMUNICATIONS COMMISSION Marlene H. Dortch Secretary 28 Federal Communications Commission FCC 02J- 2 APPENDIX A STATE COST CLUSTER ANALYSIS STATE State CPL CPL Difference Cluster Split Differences 2.50 2.00 1.55 1.50 1.00 0.85 0.50 Dist. Of Col. 16.0313 New Jersey 18.09308 2.061777043 * * * * * * California 18.36228 0.269202909 Massachusetts 19.18045 0.818170743 * New York 19.546 0.365544777 Nevada 19.71822 0.172222482 Florida 19.86697 0.148755428 Maryland 19.90223 0.035252677 Rhode Island 19.99115 0.088927016 Pennsylvania 20.64201 0.650853312 * Arizona 20.7334 0.091390437 Hawaii 20.77071 0.037317385 Illinois 20.7727 0.001985659 Utah 21.19423 0.421532696 Virginia 21.87975 0.685516771 * Texas 21.90215 0.022404339 Alaska 22.01712 0.114966706 Connecticut 22.06188 0.044759922 Georgia 22.14027 0.078388394 Minnesota 22.25597 0.11569615 Washington 22.31244 0.056473756 Wisconsin 22.71888 0.406439276 North Carolina 23.18681 0.467928123 Colorado 23.35272 0.16591541 Ohio 23.36297 0.010249713 Oregon 23.41386 0.050887279 Michigan 23.50088 0.087019363 North Dakota 23.97012 0.469238239 Indiana 24.18936 0.219238314 Iowa 24.28756 0.098201185 Kansas 24.71672 0.42916505 Puerto Rico 24.88505 0.168325153 Missouri 25.07276 0.187712385 New Hampshire 25.09483 0.022071128 New Mexico 25.7026 0.607770941 * South Carolina 26.05983 0.357227222 Tennessee 26.37502 0.315191549 Oklahoma 26.38137 0.006345897 Louisiana 26.41917 0.037807808 Idaho 26.9214 0.502222235 * South Dakota 27.77254 0.851149266 * * Arkansas 27.96557 0.193025237 Nebraska 28.20475 0.239177138 Kentucky 29.78325 1.578497802 * * * * * Maine 30.41798 0.634734615 * Alabama 31.64293 1.22494751 * * * Vermont 32.37634 0.733413484 * Montana 32.72822 0.351882006 West Virginia 33.43617 0.7079476 * Wyoming 33.71669 0.2805202 Mississippi 37.78217 4.065475787 * * * * * * * Nat. Average 21.92357 135% of Avg. 29.59681 Note: The asterisks (*) indicate which states have a Cost per Loop (CPL) difference that exceeds the cluster split difference. 29 Federal Communications Commission FCC 02J- 2 APPENDIX B PARTIES FILING INITIAL COMMENTS Commenter Abbreviation AT& T Corp. AT& T Beacon Telecommunications Advisors, LLC Beacon BellSouth Corporation BellSouth Competitive Universal Service Coalition CUSC General Communications, Inc. GCI Maine Public Service Commission Rural State Commissions Montana Public Service Commission and Vermont Public Service Commission Massachusetts Department of Telecommunications And Energy MDTE Missouri Office of the Public Counsel MOPC National Rural Telecom Association and Organization for the Promotion and and Advancement of Small Telecommunications Companies NRTA and OPASTA National Telecommunications Cooperative Association NTCA Ohio Consumers’ Counsel, Maryland Office of People’s Counsel, Maine Public Advocate Office, Texas Office of Public Utility Counsel And Pennsylvania Office of Consumer Advocate OCC et al. Qwest Communications International, Inc. Qwest Rural Iowa Independent Telephone Associaton RIITA SBC Communications, Inc. SBC Texas, Public Utility Commission Texas PUC United States Telecom Association USTA Verizon telephone companies Verizon 30 Federal Communications Commission FCC 02J- 2 APPENDIX C PARTIES FILING REPLY COMMENTS Commenter Abbreviation AT& T Corp. AT& T Florida Public Service Commission FPSC GVNW Consulting, Inc. GVNW Maine Public Service Commission Rural State Commissions Montana Public Service Commission and Vermont Public Service Commission National Telecommunications Cooperative Association NTCA Ohio Consumers’ Counsel, Maryland Office of People’s Counsel, Maine Public Advocate Office, Texas Office of Public Utility Counsel And Pennsylvania Office of Consumer Advocate OCC et al. Qwest Communications International, Inc. Qwest SBC Communications, Inc. SBC Sprint Corporation Sprint Verizon telephone companies Verizon Wyoming Public Service Commission WPSC 31 Federal Communications Commission FCC 02J- 2 2 Second, the Joint Board has responded to the court’s charge to create a state inducement mechanism to ensure that the rates actually paid by consumers are reasonably comparable. By requiring states to provide a certification and data about the comparability of urban and rural rates, the Commission ensures that ultimate rates as engineered by the state regulators reflect the same equity as the cost analysis and support provided at the federal level. In the event that state action and federal cost- based support prove insufficient, we have also created a mechanism for states to make individualized showings that modification of the cost benchmark or additional tailored support is warranted. All members of the Joint Board worked extremely hard to develop this process to ensure the continued success of the universal service support mechanism for non- rural carriers. We are all deeply committed to this goal. I look forward to working with my colleagues at the FCC and the Commission’s staff to review these recommendations and promptly implement a non- rural support mechanism consistent with the Tenth Circuit’s directives. 33 Federal Communications Commission FCC 02J- 2 STATEMENT OF COMMISSIONER MICHAEL COPPS Re: Federal State Joint Board on Universal Service, CC Docket No. 96- 45 A core principle of the Telecommunications Act of 1996 is that all Americans should have access to reasonably comparable services at reasonably comparable rates. Congress was particularly concerned about consumers in rural, insular, and high- cost areas, whether served by rural or non- rural carriers. In this Recommended Decision on a mechanism for high- cost support for non- rural carriers, the Joint Board recommends the use of statewide average costs in an initial effort to direct support to high- cost states. I support this methodology, but only insofar as there is an effective comparison of rural and urban rates at the end of the process to ensure that the statutory directive is met. The Joint Board recommends that the Commission establish a safe harbor for comparing urban and rural rates, and provide any additional funding to reach this benchmark. The Joint Board indicates that 135 percent may be the appropriate benchmark. I do not find adequate evidence in the record to demonstrate that a rural rate for telephone service that is 135 percent higher than the average urban rate in this country is necessarily or reasonably comparable. Nevertheless, I support this Recommended Decision because the Joint Board expressly requests that the Commission further develop the record to establish the appropriate benchmark. I look forward to a full record on this issue from a wide variety of stakeholders when the Commission takes up this proceeding. Finally, I emphasize that this Recommended Decision applies only to non- rural carriers serving rural America. The Joint Board has recognized that the assumptions in this decision such as using statewide average costs may well not be appropriate for rural carriers. I urge the Commission and the Joint Board to move forward expeditiously to address outstanding issues related to the high- cost mechanism for rural carriers to ensure that sufficient support is provided as required by the statute. Congress has been clear – our duty is to ensure that comparable technologies are available all across this nation at affordable and roughly equivalent rates. Each and every citizen of this great country should have access to the wonders of communications. The statute and the public interest require no less. 34 Federal Communications Commission FCC 02J- 2 SEPARATE STATEMENT OF BILLY JACK GREGG, DIRECTOR OF THE CONSUMER ADVOCATE DIVISION, PUBLIC SERVICE COMMISSION OF WEST VIRGINIA Re: Federal State Joint Board on Universal Service, CC Docket No. 96- 45 Rates matter. Congress said so in the Telecommunications Act of 1996. Now, nearly seven years after the passage of the Act, the Joint Board has taken action to ensure that the real rates that real customers pay will be the ultimate test of the success of state and federal universal service policies. Telecommunications customers throughout the nation will be the beneficiaries of this decision. In Section 254( b)( 3) of the Act, Congress established the federal universal service principle that “… consumers in all regions of the Nation, including low- income consumers and those in rural, insular and high- cost areas, should have access to telecommunications … at rates that are reasonably comparable to rates charged for similar services in urban areas.” [Emphasis added.] Although the wording of the Act is clear, up until now the Joint Board and FCC have designed universal service support mechanisms which steadfastly avoided consideration of the rates consumers actually pay. The 10 th Circuit Court of Appeals rejected the FCC’s argument that it had “… no duty to ensure the reasonable comparability of rural and urban rates… .” 136 The Court held that the FCC was “… obligated to formulate its policies so as to achieve the goal of reasonable comparability…” and “… then to assess whether its funding mechanism will be sufficient for the principle of making rural and urban rates reasonably comparable.” 137 In response to the Court’s remand, the Joint Board has finally added the critical missing piece to the existing funding mechanism. The procedures recommended by the majority in this decision will incorporate a rate review by the states as a part of the certification required under Section 254( e) of the Act. Such review should induce the states to take actions to use existing state and federal resources to achieve rates which are comparable to a national urban standard. If existing resources are inadequate, the states may request additional federal actions to enable comparable rates, including supplementary rate support. By definition, any additional federal funding will be targeted and sufficient to achieve the statutory goal of comparable rates. This common sense approach will properly respect the complementary roles of the federal and state authority in achieving rate comparability, and fairly balance the interests of high- cost and low-cost states in funding demonstrated universal service needs. 136 Qwest Corp. v. FCC, 258 F. 3d 1191 (10 th Cir. 2001) at 1200. 137 Qwest Corp. v. FCC, 258 F. 3d 1191 (10 th Cir. 2001) at 1201- 1202. 35 Federal Communications Commission FCC 02J- 2 2 The universal service principle of rate comparability contained in the Act means that regardless of the evolution and vicissitudes of the competitive telecommunications market, the rates that consumers pay for basic services in the most rural and remote areas of our Nation will stay within shouting distance of the rates paid by their urban cousins. As the dynamic forces unleashed by the introduction of competition into all areas of telecommunications continue to surprise us, universal service provides a safety net that ensures that all Americans will benefit from this revolution, even if competitive choices are not immediately available in any particular area. The procedures we recommend in this decision will go a long way toward making the universal service promise contained in the Act a reality, now and in the future. 36 Federal Communications Commission FCC 02J- 2 STATEMENT OF COMMISSIONER KEVIN J. MARTIN Approving in Part, Dissenting in Part Re: Federal State Joint Board on Universal Service, CC Docket No. 96- 45 I wish to thank all my colleagues on the Federal- State Joint Board for their hard work and contributions in the effort to reach consensus on the important issue of establishing a universal service support system for non- rural carriers. I believe that today’s effort, however, falls short in meeting our obligation to ensure that consumers living in rural and high cost areas have access to similar telecommunications services at rates that are reasonably comparable to rates paid by urban consumers. Congress gave the Commission a clear mandate: to ensure that consumers in all regions of the nation have access to services that “… are reasonably comparable to those services provided in urban areas and that are available at rates that are reasonably comparable to rates charged for similar services in urban areas (emphasis added).” 138 Congress’ direction is also clear regarding the obligation to establish mechanisms that are “… specific, predictable and sufficient… to preserve and advance universal service.” 139 In remanding the Commission’s previous attempt to establish a federal- high cost universal service support mechanism for non- rural carriers, the United States Court of Appeals for the Tenth Circuit agreed that these fundamental guiding principles govern Commission action on any policies regarding universal service support mechanisms. 140 Despite this remand, the majority’s recommendation essentially reaffirms the Commission’s existing universal service support mechanism for non- rural carriers. The decision continues to base support on forward looking costs and creates a sparsely defined second supplemental support system based on rate comparisons. Today’s recommendation falls short in its response to the court mandate that we define the statutory term “reasonably comparable” for purposes of the cost- based support mechanism and fails to demonstrate, with any degree of specificity, how the proposed secondary mechanism will satisfy the statutory requirement that universal service support be “specific, predictable and sufficient.” For these and the reasons explained below, I respectfully dissent from portions of the majority 138 See 47 U. S. C. 254( b)( 3). 139 See 47 U. S. C. 254( b)( 5). 140 Federal- State Join Board on Universal Service, CC Docket No. 96- 45, Ninth Report and Order and Eighteenth Order on Reconsideration, 14 FCC Rcd 20432( 1999)( Ninth Report and Order) remanded, Qwest Corp. v. FCC, 258 F. 3d 1191, 1199 (10 th Cir. 2001). 37 Federal Communications Commission FCC 02J- 2 2 opinion today. 141 Use of Costs as a Surrogate for Rates to Determine Non- Rural High Cost Support Section 254( b)( 3) of the Communications Act requires that universal service support mechanism ensure that telecommunications services in all regions of the nation be provided at reasonably comparable “rates.” The majority, however, recommends continuing the practice of using costs rather then rates to determine federal support. I am not convinced that a mechanism based solely on costs would meet the statutory mandate requiring a comparison of rates. Moreover, I fear that the recommended decision may be either arbitrary or not fully thought through. If the Joint Board is confident that a cost- based support system satisfies our statutory obligation to produce reasonably comparable rates, then why does it propose establishing an entirely new support mechanism based on rate comparisons? Similarly, if the Commission were to adopt the Joint Board’s recommended “supplemental rate comparability review,” why should it not abandon the cost- based support mechanism and instead rely solely on the rate- based support mechanism? If we need the supplemental rate comparison to meet our statutory obligation, would it not be simpler to have only one mechanism rather than two? These questions seem to remain unanswered by the majority. The majority’s rejection of rate- based distribution and support for a cost- based mechanism is based on two arguments: (i) an analysis of disparate local rate design practices throughout the nation remains too difficult a task; and (ii) the use of costs reflects the federal government’s primary obligation to support only those states that “do not have the resources within their borders to support all of their high cost lines.” 142 In my view, both of these arguments fail to support the Joint Board’s position. First, in response to the argument that such an analysis is too difficult, the majority appears to create just such an analysis in its “supplemental rate comparability review.” The majority also fails to note or even address the fact that many of the issues and data necessary to perform a rate-based comparison will be needed in the context of initiating the proposed catch all “supplemental rate comparability review.” On its face, one well- defined support mechanism based on rate comparisons would appear to present an equal or lesser administrative burden for the Commission, the states and carriers compared to the dual cost- based and rate- based mechanisms recommended by the majority today. The majority’s recommendation also contains an inherent presumption that the federal government’s role in establishing a support mechanism is apparently limited to equalizing cost discrepancies between states but not equalizing rate discrepancies between rural and urban 141 In addition to the reasons discussed below, I also agree with and join in many of the concerns raised in Commissioner Bob Rowe’s thorough and thoughtful analysis in his dissent. 142 Recommended Decision at paras. 19- 21, 24. 38 Federal Communications Commission FCC 02J- 2 3 areas. 143 I disagree. The statute is clear. Our job is to ensure that services in rural and high cost areas are “available at rates that are reasonably comparable to rates charged for similar services in urban areas.” 144 The 10 th Circuit explicitly rejected the FCC’s contention that it had no duty to ensure the reasonable comparability of rural and urban rates and stated that we are “obligated to formulate policies so as to achieve the goal of reasonable comparability…” 145 In my view, if the Commission is only going to address discrepancies between and among states, then there must be a requirement that states address such discrepancies within their borders. Whether such a requirement compels rate averaging within states or requires that a state universal service mechanism be in place, such action must address differences in cost between rural and urban areas. Yet this decision fails to require that such inequities between urban and rural rates be addressed. 146 The proposed expanded rate certification mechanism is insufficient. Under the proposed certification process, states would be permitted to report rates that are not “reasonably comparable” according to the benchmark. Such rates could eventually be allowed to meet the “reasonably comparable” standard if a state demonstrates “additional services included in the basic service rate” or by outlining “the method in which the state has targeted existing universal service support.” 147 In my view, such a certification process is insufficient without a standard enunciating the allowable discrepancy for intrastate rates. 143 Recommended Decision at paras. 25- 26. “The Commission’s primary role is to identify those states that do not have the resources within their borders to support all of their high- cost lines.… The Commission explained in the Ninth Report and Order that the non- rural high cost support mechanism “has the effect of shifting money from relatively low- cost states to relatively high- cost states. The Commission believed that its non- rural support mechanism ensured that no state with costs greater than the national benchmark would be forced to keep rates reasonably comparable without the benefit of federal support…. We continue to support these policies.” 144 47 U. S. C. 254( 3). 145 258 F. 3d at 1200. 146 See Ninth Report and Order at 20482- 3, para. 95 (The Commission found it most appropriate to allow states to determine how non- rural cost support is used, “[ b] ecause the support… is intended to enable the reasonable comparability of intrastate rates, and states have primary jurisdiction over intrastate rates.”; see id. At 20483, para. 96 (“ As long as the uses prescribed by the state are consistent with 254( e), we believe that states should have the flexibility to decide how carriers use support provided by the federal mechanism.”). See Recommended Decision at paras. 43- 56. Even in light of the 10 th Circuit remand requiring the Commission to consider appropriate state inducements to address reasonably comparable rates, the Joint Board fails to consider recommending either a state averaging mandate or mandatory state universal service mechanism requirement to address discrepancies between costs in rural and urban areas. 147 Recommended Decision at para. 55. 39 Federal Communications Commission FCC 02J- 2 4 Sufficiency of High- Cost Support under the National Average Cost Benchmark Even if costs can be used as a surrogate, I question the majority’s recommendation to use the 135% benchmark to ensure that rural rates are “reasonably comparable.” In deciding to proceed with a cost- based methodology to ensure reasonably comparable “rural” and “urban” rates, we should compare “rural” costs to average “urban” costs. The Commission certainly has data readily available to perform this comparison. Under the Synthesis Cost model, cost data can be produced by density zone or at the wire center level. Yet, the majority summarily rejects the concept of an “urban benchmark,” setting a benchmark at 135 percent of national average cost. In the process, the decision sidesteps the question of whether the benchmark produces sufficient support in light of the existing disparity between national average cost and the lower average urban cost. As Commissioner Rowe notes, the majority’s rejection of the urban benchmark is “confusing and unpersuasive.” 148 The majority never tackles the uncomfortable fact that the 135 percent benchmark is too high because national average costs are already higher than urban costs because they include in the national average the very rural areas at issue. In other words, the high costs associated with serving rural areas are used twice: once to raise the national average and again in comparison. Let me illustrate my concern with a simple example. If half of the country lived in an urban area with costs of $10 and the other half of the country lived in a rural area with costs of $30, the difference between the costs of the average urban area and average rural area would be $20. But if a national average were taken, including the costs of the rural areas, the national average cost would be $20. If support were then based on the difference between the rural cost ($ 30) and 135% of the national average (1.35 x $20=$ 27), each rural resident would have costs of $27 ($ 30-$ 3 of support) and each urban resident would have costs of $13 ($ 10 + $3 of support). I do not believe that such a methodology sufficiently addresses the reasonable comparability of rural and urban costs. The inclusion of rural costs in the average along with the adoption of a 135% benchmark systematically underestimates the costs of rural areas. 149 Instead, the majority finds fault with the use of an urban benchmark based on the fact that it “substitutes costs for rates” and “compares statewide average costs to nationwide urban costs.” 150 The majority’s criticism appears strangely out of place given that its own recommendation is also based on a cost- based support system. I find it ironic that the majority can justify its “existing system on the ground that costs equal rates, and at the same time rejects all changes on 148 See Commissioner Bob Rowe’s Separate Statement at 8. 149 National averages could be used without a benchmark or urban averages could be used with a benchmark but the combination of the two mechanisms is arbitrary. 150 Recommended Decision at para. 39. 40 Federal Communications Commission FCC 02J- 2 5 the ground that costs do not equal rates.” 151 It also rejects the urban benchmark because it would “require more funding or a higher benchmark level because urban average costs are lower than national average costs.” 152 I fail to see how the potential for greater funding levels should prevent us from adopting a support system that meets our statutory obligation. 153 Indeed, I fear that this reasoning reflects an analysis that concluded first that there would be no additional funding for rural areas and second adopted a mechanism to assess “reasonable comparability” that achieved that result. I believe our statutory obligation was to achieve reasonably comparable urban and rural rates even if that “requires more funding” than the current system provides. Nor do I understand how the majority reaches the conclusion that the urban cost benchmark fails to “better satisfy the statutory comparison of urban and rural rates.” 154 I join Commissioner Rowe in questioning how the majority finds that additional “incremental support would be ineffective at producing comparable rates, but existing support passes the test.” In addition, I question the use of forward- looking costs as the basis for distributing universal service support. Today, rates are set in most states through the use of actual costs not hypothetical replacement costs. Forward- looking costs have little, if any, nexus to the establishment of end user retail rates. Use of these costs for calculating universal service support results in support being provided to some areas with low end user rates while certain areas that have high rates receive insufficient support. In my view, we could better achieve comparability of rates if we based our universal service support system on actual rather than forward looking costs. Finally, the majority cites three studies/ analyses in support of its decision to continue using the 135 percent benchmark. I disagree with the majority’s conclusion that these studies support its decision to retain the benchmark. First, the majority points to the General Accounting Office (GAO) study to show that national averages of rural, suburban and urban rates are affordable and reasonably comparable. The majority, however, fails to acknowledge serious deficiencies in the GAO study that fail to support the use of the benchmark for non- rural carriers. 155 For example, the GAO study includes data from areas served by rural carriers, areas that are not relevant to the 151 See Commissioner Bob Rowe’s Separate Statement at 8. 152 Id. at 40. 153 The United States Department of Agriculture’s Rural Utilities Service (RUS) recommended adoption of a benchmark tied to the national average urban loop cost or another statistical indicator more representative of urban costs, not the national average costs. RUS notes that 135% of the national average (urban and rural) “loop cost” exceed its estimate of urban “loop costs” by 233%. 154 Recommended Decision at 39. (emphasis in original). 155 See also Commissioner Rowe’s Separate Statement at 2- 3. 41 Federal Communications Commission FCC 02J- 2 6 establishment of non- rural carrier support system. In addition, GAO’s rate comparison ignores whether rates in different service areas apply to comparable services. Moreover, national averages cited by GAO do not assist the Commission in addressing our core responsibility of whether rates in certain rural or high cost areas are comparable to rates in urban areas, or even whether rates vary significantly from state- to- state. To the contrary, as Commissioner Rowe points out, GAO’s data demonstrates a vast disparity on state rates (e. g., residential rates at two Wyoming locations exceeding $40 versus residential rates in Roaring Springs, Texas of $7.10). 156 I also join Commissioner Rowe’s dissent asserting that a standard deviation analysis fails to justify the current benchmark. 157 I find it particularly troubling that the majority arbitrarily raises the benchmark to 135 percent even in light of its own analysis demonstrating that 2.0 standard deviation above the national mean results in a 132 percent benchmark. The majority offers no reasoned basis why states should be denied the additional $. 50 per customer per month of support that would result by applying the results of the majority’s own standard deviation analysis. Supplementary Rate Review The majority, in today’s recommendation, sets forth an additional supplemental process for rate comparison. It recommends adopting a new and vaguely defined supplemental mechanism. Rather than provide a clearly defined mechanism the majority instead offers an ad hoc process where the specific mechanisms will apparently develop on a case- by- case basis. 158 The majority envisions a process where States seeking additional federal support will be required to provide a “rate analysis,” and will have “great flexibility” in demonstrating that rates are not reasonably comparable. 159 In my view, the majority’s “supplementary rate review” is striking similar to the state- by- state cost study approach the Commission had originally rejected in order to pursue its flawed nation-wide universal service cost model approach. Under the recommended state- by- state approach, each state would have significant latitude to suggest its own procedures for adjusting rates. Without specific guidelines or a clearly defined standard, this approach appears to invite the potential for uneven and potentially discriminatory results. I am troubled that majority fails to offer any specific guidance on critical areas of its newly proposed process. The item is silent, for example, on whether states should alter rates to take into account the scope of certain local calling areas or differing calling plans. In my view, 156 Id. at 3. 157 Id. at 5- 7. 158 Id. at 16. 159 Recommended Decision at para. 56. 42 Federal Communications Commission FCC 02J- 2 7 without an established standard or guidance for states in this area, the poorly defined “supplementary rate review” will most likely provide results, if any, that are highly susceptible to legal challenge. Finally, Commissioner Rowe is correct in questioning whether the proposed “supplementary rate review” would “create perverse incentives for carriers.” 160 One of the reasons the Commission adopted the forward- looking cost model was because it believed that an embedded- cost support system promotes inefficient investment that would inhibit competitive entry. I find it ironic that the majority now seeks to adopt a rate- based mechanism that inherently relies on local rates which are typically based on embedded costs. 160 See Commissioner Bob Rowe’s Separate Statement at 18. 43 Federal Communications Commission FCC 02J- 2 SEPARATE STATEMENT OF COMMISSIONER THOMAS J. DUNLEAVY NEW YORK STATE PUBLIC SERVICE COMMISSION Approving In Part and Dissenting In Part Re: Federal State Joint Board on Universal Service, CC Docket No. 96- 45 I support this Recommended Decision (RD) to the extent that it confirms and endorses continued use of the Commission’s existing high- cost support mechanism for non- rural carriers. I agree with the Joint Board’s suggestion that existing rates are affordable and reasonably comparable. I have seen no evidence that services and rates in “all regions of the Nation” are not reasonably comparable. In my view, this confirms that the existing mechanism, in concert with the other available federal universal service support mechanisms, provides sufficient support to achieve the purpose for which it is intended. I believe that the standard deviation and cluster analyses cited in the RD provide sufficient justification for the retention of both the existing cost- based mechanism and its cost benchmark of 135% of the national average forward- looking cost. Further, I agree that requiring states to certify whether their rates are reasonably comparable responds adequately, as I read it, to the Tenth Circuit's requirement that the Commission somehow induce states to fulfill their own obligations to ensure reasonably comparable rates. I must respectfully dissent, however, from those portions of the RD that suggest that states should be invited to seek additional “targeted” support based on a comparison of rates and that a rate benchmark of 135% of the average urban rate may be appropriate for making that comparison. It does not appear to me that this suggestion is supported by the evidence. Because the RD itself concludes that the record for choosing a rate benchmark must be further developed, it should not suggest that any particular value might be appropriate. Having endorsed continued use of the current cost- based mechanism for calculating non-rural support, the Joint Board then incongruously suggests that states should be invited to seek supplemental support based on an analysis of their rates. In so doing, it proposes creation of two fundamentally different support mechanisms, between which each recipient state would be invited to choose to use the more lucrative for itself. The suggestion that such a supplemental support mechanism should be created implies that the majority questions the sufficiency of its primary cost- based mechanism. I do not. As explained in the RD, it is extremely difficult, if not impossible, to make meaningful rate comparisons among states. The RD hints at some of the difficulties in paragraph 19, but ignores what, in my view, is perhaps the largest problem - normalizing rates for the varying local calling capabilities they may encompass. For this reason, the Joint Board has now twice concluded that cost analysis, as a proxy for rate analysis, is the preferred approach and the Commission has so agreed. Nevertheless, the RD implies that the Commission will indeed be able to create the necessary rules and algorithms to perform such normalizations, make such rate comparisons and, most importantly, determine appropriate amounts of supplemental support. With the greatest respect and admiration for the skills of the Commission, I do not share the majority’s optimism on this score. 44 Federal Communications Commission FCC 02J- 2 2 Additionally, I am also concerned that the RD may contemplate comparison of each wire center's rate to a national benchmark rate in order to calculate the proposed supplemental rate-based support. If so, it seems to me that it would reverse its own finding that high- cost support should be calculated based on statewide, not wire center, costs. That would fundamentally alter the Commission’s role in high cost support, which as both the Joint Board and the Commission have found, is to effect necessary state- to- state transfers of monies, not transfers within states. It appears to me that, in the final analysis, the uncertainty that surrounds this ill- defined support proposal is its most troubling attribute. Until the extremely difficult decisions have been made about how to normalize rates and what criteria to apply in determining supplemental support, it is impossible to even estimate how much support it might produce or where that support might go. Potential recipients cannot even guess how much support they ultimately will receive; payers can only speculate on how much cost they will be asked to bear. Consequently, the Commission will be hard pressed, in my view, to explain to the court how its mechanism will, in fact, produce reasonably comparable rates. 45 Federal Communications Commission FCC 02J- 2 SEPARATE STATEMENT OFCOMMISSIONER BOB ROWE, MONTANA PUBLIC SERVICE COMMISSION Dissenting Re: Federal State Joint Board on Universal Service, CC Docket No. 96- 45 I respectfully but strongly dissent from this Recommended Decision. The 10 th Circuit Court of Appeals required the FCC to demonstrate that the statutory requirements of section 254 of the Act were met. In order to do this, the FCC was required to define and recognize in its decision certain key statutory terms, such as “reasonably comparable” which animate section 254. The court also directed the Commission either to explain how the 135 percent benchmark used in the current formula complies with the established definition of “reasonably comparable” or to establish a new benchmark that does meet the statutory requirements, including the requirement of providing sufficient support to allow states to achieve reasonably comparable rates. The Recommended Decision fails these tasks. The Recommended Decision does not demonstrate that the Joint Board has impartially examined and addressed these questions. The recommended decision does not even demonstrate that the Joint Board has considered these issues in a manner that I believe was contemplated by the Court. Rather, the Recommended Decision embraces a variety of flawed arguments that purport to justify continuation of the status quo. I am unpersuaded by these arguments and conclude that if the Recommended Decision were adopted by the Commission some “nonrural” carriers (more properly large companies serving high cost areas) would not receive sufficient support, and rural rates and services in some states could not be reasonably comparable to those in urban areas. The majority concludes that the statute can be satisfied by two quite different support systems: continuation of the existing system based on forward- looking costs; and a second, supplemental, support system based on rates. For different reasons, I conclude that each is seriously flawed. Even taken together, they fail to provide sufficient support to meet the legal requirements of the Act. A. Cost- Based Support The majority recommends continuing primary reliance on the existing cost- based support system, and continuation of the current 135 percent benchmark. That benchmark is a critically important variable in the support system. The benchmark is now set by rule at 135 percent of the national average cost produced by the Commission’s forward-looking cost model. 161 In 2002, that works out to $29.60 per line per month. 162 161 47 C. F. R. § 54.309( a)( 3). 162 This is 135 percent of the current national average cost among nonrural carriers of $21.9235. 46 Federal Communications Commission FCC 02J- 2 2 This “dollar benchmark” has a simple meaning for nonrural carriers: in any state where the average unseparated forward- looking cost of providing service is more than $29.60, federal support will reduce the net cost in that state to $29.60. This is accomplished by a combination of cost separation policy 163 and universal service support. 164 A dollar benchmark that is too high produces two forms of harm. In the very highest cost states, too little support is provided. The net costs of high- cost carriers in those states necessarily remains too high, above a level reasonably comparable to the net costs of carriers serving urban areas. Second, states with somewhat lower cost receive no support whatever, even though their costs are above the level that is reasonably comparable to costs in urban areas. The majority offers three justifications for keeping the 135 percent benchmark: a study by the United States General Accounting Office (GAO), “cluster analysis,” and “standard deviation analysis.” In my view, each of these justifications has fundamental flaws, and none of them support the majority’s conclusion that the 135 percent benchmark should be retained. Moreover, the Recommended Decision incorrectly rejects arguments for an “urban benchmark” and overlooks substantial evidence that the existing benchmark produces insufficient support to nonrural carriers. Finally, the majority never defines “reasonably comparable” in reference to cost- based support as is required by the Act and the 10 th Circuit’s decision. 1. The GAO Study The Recommended Decision cites a study of telephone rates conducted by the GAO. The GAO gathered data from state commissions on local telephone rates in sampled locations throughout all fifty states and the District of Columbia. 165 Within each state, GAO randomly selected locations from three broad categories associated with population density: central city, suburban (other places within a metropolitan statistical area, MSA), and rural (outside MSA). Based on data contained in the GAO study, the 163 Some of the operating and capital costs of each incumbent local telephone company are “separated” to the interstate jurisdiction. Revenues to recover these costs come from a variety of sources controlled by the Commission. 164 High cost support amounts to 76 percent of the difference between the state’s average forward- looking cost and $29.48. The 76 percent figure was selected as the average separations factor nationwide. It was adopted to prevent double- recovery of costs that are otherwise recovered through interstate revenues. 165 For residential and single- line business customers, GAO asked for the unlimited service rate and the message or measured service rate with the lowest rate. The rates do not include the Federal Subscriber Line Charge; state and local surcharges for items such as state universal service funding, 911 service, and taxes; the federal excise tax; or long distance fees and associated universal service charges and other taxes. Where offered, GAO used the tariff rate for unlimited service. Where unlimited service is not available, GAO calculated a monthly fee for a “representative customer” using the message rate. GAO assumed that a “representative customer” makes 100 5- minute calls per month. 47 Federal Communications Commission FCC 02J- 2 3 majority finds that “the national average of rural, suburban and urban rates for residential customers diverge by less than two percent.” While this may be a correct statement, based on their methodology, for three reasons it fails to support the 135 percent benchmark. First, the GAO study collected data from all parts of the nation, including areas served by nonrural carriers and areas served by rural carriers. The GAO made no attempt to exclude data from areas served by rural carriers, areas that are not under consideration here. 166 The support systems for rural and nonrural carriers are different in important ways, and data from rural carriers can mask important trends in the rates of nonrural carriers. A related deficiency is that the GAO made no effort to determine whether the rates that they measured in different areas applied to comparable services. Therefore the GAO failed to control for two significant variables, and its study conclusions cannot inform the selection of a benchmark for nonrural carriers. A more fundamental problem is that national averages are at the wrong geographic scale to inform the benchmark issue. The national average is defined by low-rate areas as well as high- rate areas. This overlooks the relevant policy variable, which is state- to- state rate variation. The question under section 254 is whether rates or costs in one or more rural, insular or high cost areas are too high in relation to rates or costs for comparable services in urban areas. National averages are irrelevant to this question. 167 It is easy to imagine a situation in which there is a section 254 problem but national averages are exactly equal. The GAO’s own data demonstrate that this is not a hypothetical problem and that rates vary significantly from state to state. According to the report, residential rates in Roaring Springs, Texas were $7.10 per month. By contrast, all of Wyoming and all of Vermont reported residential rates above $23. At the extreme, residential rates at two locations in Wyoming reportedly exceeded $40. In short, the GAO data tend to show, if anything, that there are rate differences in the country that are significant enough to be of concern under section 254. Finally, the GAO Report itself had several methodological problems. The sample size used in the study was too small to be statistically valid for any state. Second, the study overlooked some kinds of local exchange charges that must be paid by all end users. Third, the study reported raw rates, and made no correction for calling area size or other quality of service variables. Finally, in areas where more than one local calling 166 The GAO did not develop consistent data from state to state. In some states rural data was included, but not in other states. 167 The scale problem can be understood more generally with a simpler set of facts. Suppose that every state has geographically averaged rates. (This much is not far from the truth in many states.) Further suppose that in 40 states all customers pay $5 per month and in the other 10 states all customers pay $50 per month. In each state considered separately, the GAO study would show that rates for rural and urban areas are equal. Therefore when all states are aggregated together, urban and rural averages will also be equal. This is true even though rates in some states are ten times the rates in most states. 48 Federal Communications Commission FCC 02J- 2 4 area is offered, the GAO did not even pick consistent rates. For example, in Michigan the study reported the local rate as $49 per month, but a footnote shows that the most common rate is $12 The GAO study was a commendable effort by an agency that has no expertise in the complexities of local exchange rates. I do not believe that the GAO data have been shown to be sufficiently accurate and normalized so that they can support relevant inferences here. 2. Cluster Analysis The majority also found that the 135 percent benchmark is empirically supported by “cluster analysis.” This is described as “an analytical technique that seeks to organize information about variables so that relatively homogeneous groups, or clusters can be identified.” The underlying data are the statewide average costs produced by the Commission’s forward- looking cost model. Cluster analysis places the states in rank order of cost and then examines the cost differences between successive states. Where a significant cost difference exists from one state to its next neighbor in the list, cluster analysis considers this a significant fact. The highest cost state is Mississippi. This year, Mississippi and the next seven states on the ranked list receive support. The last of these, Kentucky, has a cost of $29.78. The very next state on the list, which receives no support, is Nebraska. It has a cost of $28.20. The cost difference between Kentucky and Nebraska is therefore $1.58. As it happens, this is an unusually large gap; most other states are separated from their neighbors by smaller cost differences and thus are more like their neighbor in the list than are Nebraska and Kentucky. The majority concludes that cluster analysis shows that the “135 percent benchmark targets support to states with substantially higher average costs than other states.” 168 Of course, any benchmark would target support to the highest cost states. That is nothing more than a design feature of the support system. Therefore, the majority must be placing reliance on the size of the cost difference between Kentucky and Nebraska. The $1.58 gap between Kentucky and Nebraska is therefore the sole fact underlying the majority’s argument. Cluster analysis does not provide any useful information or guidance in selecting a specific benchmark. The method used to select the clusters is arbitrary, and the approach suffers from several deficiencies. Most basically, it says nothing about whether support is sufficient. While the analysis does show that states like Kentucky (and higher cost states) are at least somewhat different from states like Nebraska (and lower cost states), it says nothing about whether costs in the higher costs states (or in Nebraska for that matter) are reasonably comparable to urban areas. Second, even assuming for the sake of argument that cluster analysis could have some probative value, the majority has justified a permanent feature of the support 168 Recommended Decision, ¶ 34. 49 Federal Communications Commission FCC 02J- 2 5 system on transient data. The 135 percent benchmark is a fixed parameter that is permanently codified in rule. Yet each year that fixed benchmark is applied to a new set of cost data, and the individual state costs change significantly over time. 169 Between the 2001 and 2002 support years, more than two- thirds of the states saw their average cost change by $0.50 or more 170 and one- third saw their average cost change by $1.00 or more. 171 When data are that volatile from year to year, the results from any one specific year, such as 2002, cannot fairly support the selection of a permanent benchmark. The effect of data volatility is illustrated by applying cluster analysis to data from other years. In 2001, for example, the cost gap between Nebraska and Kentucky was even larger than in 2002. 172 Cluster analysis would presumably have concluded even more strongly that Nebraska’s lower- cost group and Kentucky’s higher- cost group should be treated differently. However, in 2001, as it happened, neither Kentucky nor Nebraska received support. If cluster analysis had been applied a year ago, then, it would have supported reducing the 135 percent benchmark so that Kentucky could get some support. Third, the cluster analysis does not provide a unique solution at the 135 percent benchmark. Cluster analysis purportedly shows that Kentucky (and higher cost states) should receive support and that Nebraska (and lower cost states) should not. However, that result can be achieved by selecting a benchmark as low as 129 percent and as high as 135.5 percent. The majority does not explain how cluster analysis guided its selection of 135 percent from this range. This is an important distinction because the use of a 129 percent benchmark would have increased support to every state now receiving support by $1.00 per line per month. 173 In summary, the majority uses cluster analysis to support retention of the existing 135 percent benchmark even though cluster analysis is irrelevant to sufficiency, even though the method attributes inappropriate significance to apparently transient cost differences, and even though the analysis equally supports the adoption of significantly different benchmarks. 3. Standard Deviation Analysis The majority asserts that the 135 percent benchmark is supported by “standard deviation analysis” since “an analysis of the Commission’s cost model shows that two 169 Between 2001 and 2002, costs in 18 states declined by more than $1.00 per line per month, and costs in three states declined by more than $2.00. 170 This group comprised 36 of 52 jurisdictions. 171 This group comprised 18 of 52 jurisdictions. 172 The gap in that year was $2.12. 173 A benchmark of 129 percent is 6 percent lower than 135 percent. The support difference for each carrier already receiving support would be $21.93 x 6% x 76% = $1.00. 50 Federal Communications Commission FCC 02J- 2 6 standard deviations translates approximately to a 135 percent cost benchmark.” 174 The majority explains that standard deviation analysis is a “commonly used statistical analysis” and that using two standard deviations will “identify data points which are truly outliers within the sample studied.” 175 I have concluded that standard deviation analysis does not validate the 135 percent benchmark. The most fundamental problem is that the majority would substitute for the existing statutory standard a new and different standard that is not even conceptually related to reasonable comparability. 176 The standard deviation analysis apparently relies on the premise that only “outlier” states should receive support. But this measure of who is an “outlier” is defined by standard deviations. This in turn depends not only on the magnitude of the cost differences between a high- cost state and urban areas, but by the cost assigned to every state. The overall pattern of state costs, rather than the differences between the extremes defines the outcome. Even where urban costs are low, the presence of many semi- rural states would raise the national average and could also increase the standard deviation. Each of these forces would reduce support for high- cost states. Indeed, it is easy to imagine circumstances under which no state is more than two standard deviations above the mean, but many have costs that are not by any reasonable measure comparable to urban costs. The majority has not shown that, or even discussed whether, the two standards are the same. The absence of that showing has led to the second flaw with this approach. Standard deviation analysis actually does produce a benchmark that is too high to produce sufficient support. To understand why, it is instructive to consider how far “out” a state must lie in order to be an “outlier” under the majority’s analysis. The majority reports that cases “within two standard deviations of the mean will comprise approximately 95 percent of all data points.” This is accurate under some conditions, but not here. Here we should be interested in how many high- cost cases exist, not in the total of high- cost and low- cost states. 177 Under those circumstances, only 2.3 percent of the cases in a normal distribution fall outside the limit of two standard deviations. 178 This is an important finding. If state costs were normally distributed, standard deviation analysis would suggest that support should be provided only to states in the 99 th and 98 th percentiles of cost, and to some of the states in the 97 th percentile. 179 As a result, several 174 Recommended Decision, ¶ 35. 175 Recommended Decision, ¶ 36. 176 I note that the court has already rejected earlier efforts to redefine the statutory standard. 177 This calls for use of what is normally referred to as a “one- tailed” test. 178 In a “z” distribution, the probability of a case falling between 0 and 2 standard deviations is 0.4772. This means that the probability of a case falling above 2 standard deviations is 1 – 0.5 – 0.4772 = 0.0228 or 2.28 percent. 179 Since 2.0 standard deviations includes 2.3 percent of a normal distribution, only 30% of the states in the 97 th percentile would be included. 51 Federal Communications Commission FCC 02J- 2 7 states fall below two standard deviations and below the 135% benchmark, but they are left with net costs that are still far above urban costs. This is solely the result of the fact, based on the cost patterns of all of the states, that their costs are not high enough to make them “outliers.” 180 I know of no basis to conclude that this is a reasonable interpretation of section 254. Indeed, it is a substitution of a significantly more restrictive standard. Nothing I am aware of in the legislative history of the Act suggests that Congress intended universal service support to be available only to carriers or states in the top three percentiles of cost. I conclude that standard deviation analysis would improperly substitute a new and more restrictive standard for the standard contained in the statute. It is not merely an interpretation of that statutory standard. The majority suggests the use of the standard deviation places the 135 percent benchmark on a stronger scientific footing. I do not agree. Although the analysis cites a statistical variable, there is nothing in the science and art of statistics that supports this variable as a yardstick for cost benchmarks. Standard deviation analysis is certainly useful in some scientific and public policy applications. Sometimes data will be excluded if they fall outside a statistical limit such as two or three standard deviations. The assumption is that these “outlying” measurements were the product of a measurement error of unknown origin. By excluding the suspect data, one can expect to increase the reliability of the inferences drawn from the surviving data. This procedure for excluding data is not only used in science, but, as the majority notes, it is also used in telecommunications regulation. For example, it is used to discount service quality failures that might unjustly lead to penalties unless they are disregarded. The fact that standard deviation analysis may be useful in some public policy situations does not establish that it is useful here. Here the majority apparently conflates “outlying” and unreliable data with “outlying” high- cost states. Despite the superficial similarities, I believe the two situations are fundamentally different, and the analogy is inappropriate. We are clearly not seeking here to identify any unreliable data in order to discard it. Indeed, the majority’s analysis would not lead to exclusion of any data points. On the contrary, once the benchmark has been set, the Commission will then rely on precisely the same cost data for “outlying” states to calculate and distribute tens of millions of dollars of support. In what way, then, does standard deviation analysis illuminate the task before us, which is better to define what Congress meant by “reasonably comparable?” It appears that the majority believes that we should define when rates are reasonably comparable using the same standard by which statisticians eliminate unreliable data. I know of no 180 Moreover, because the benchmark is too high, even in those states that do receive some support, the amount is insufficient. 52 Federal Communications Commission FCC 02J- 2 8 statistical theory that suggests this equation. 181 The tasks are fundamentally different, and solutions used to solve one kind of problem are not appropriate to the other. As used, standard deviation analysis does not add any statistical or scientific validity to the majority’s decision about the appropriate benchmark. Standard deviation analysis suffers from several additional problems. The majority’s analysis perpetuates one of the errors already identified by the court. It is based upon the cost characteristics of all states, not the cost characteristics of urban areas, or even of urban states (assuming such states could be found). Using standard deviation analysis based on the national sample once again simply ignores the important difference between the average costs of urban areas and the average cost of the country as a whole. Even if one were to accept the validity of standard deviation analysis for setting the benchmark, at the very least section 254 requires that analysis to be applied solely to the statistical cost characteristics of urban areas. Another problem is that the majority has inexplicably rounded up the results of its own standard deviation analysis. The record shows that a point 2.0 standard deviations above the national mean produces a 132 percent benchmark. The majority has not explained why this evidence supports a benchmark of 135 percent. Raising the benchmark by three points to 135 percent further constricts eligibility for support. The support reduction in states receiving support is $0.50 per line per month, an amount that is not trivial to a high- cost customer. 182 Finally, note my disagreement with two other statements in the Recommended Decision. I am not aware that the Joint Board has previously found that rates in all parts of the country are reasonably comparable. 183 Also, I do not agree that the major purpose of federal support is to “ensure that rates remain reasonably comparable as competition develops. 184 If rates or costs were not reasonably comparable on the day the 1996 Act 181 Nor is the majority here using standard deviation analysis to test any statistical hypothesis. Scientists sometimes use standard deviations to test a hypothesis that a sample of cases is statistically different from a larger population. No such hypothesis has been stated here, and there is no sample to be tested. 182 At a benchmark of 132 percent of the national average, every currently supported customer would receive additional support of $0.50 per customer per month. $0.50 = ($ 21.93 * (135%- 132%) * 76%) 183 The majority states that the Joint Board has previously found “in prior rulings that current rates are affordable and reasonably comparable.” Recommended Decision, ¶ 34. The former is a correct statement, but not the latter. The majority cites the Commission’s Seventh Report and Order. That document merely recited an earlier finding that rates are affordable. Nothing in the cited paragraph establishes that rates are or were reasonably comparable. 184 The majority states that this is “one of the goals” of the Act. Recommended Decision, ¶ 35. The majority’s own rationale on this issue is inconsistent. On one hand the majority apparently supports Verizon’s argument that rates were reasonably comparable in 1996 and that federal support is needed, if at all, primarily for the intrastate rate variations that were expected to follow the arrival of local competition. ¶ 35. On the other hand, the majority admits that there is no need for federal support when rates are de-averaged in a low- cost state, since with a sufficient intrastate fund all customers in the state can still have reasonably comparable rates without federal support. ¶ 26. 53 Federal Communications Commission FCC 02J- 2 9 went into effect, that is a problem under section 254. I believe the Joint Board should state clearly that the Commission has a responsibility under law to address rate and cost differences, without regard to their vintage. Section 254 creates a duty to keep rates and service reasonably comparable regardless of whether local competition has flourished or languished and regardless of whether rates have remained static or have been geographically de- averaged in response to competition. 4. The Urban Benchmark The majority definitively rejects the use of an “urban benchmark.” I find the arguments confusing and unpersuasive. Ultimately, the majority has failed to address the central issue remanded by the court, the relationship between the dollar benchmark, which represents each state’s post- support net cost, and the typical cost in urban areas. By insisting upon stating the benchmark as a multiple of national average costs, the majority seeks unsuccessfully to avoid the fact that, because urban costs are lower than the national average cost, the benchmark set at 135 percent of the national average is too high. First, the majority suggests that the distinction between rates and costs, combined with the Tenth Circuit decision, somehow prohibit us from increasing cost- based support. The majority states that the “litmus test for non- rural high- cost support is the reasonable comparability of urban and rural rates,” and that “rates do not necessarily equate to costs.” 185 Similarly, the majority states that adopting a cost benchmark that is reasonably comparable to urban cost would be to “substitute a different standard for the statutory rate comparison.” 186 I fail to understand this argument, given the majority’s own reliance on a different cost benchmark less favorable to rural customers. The majority recommends continued primary reliance on cost- based support to comply with section 254. If the majority believes that local exchange rates are the only valid basis for support, then I do not understand why hundreds of millions of dollars of support should continue to be based on costs nor why the Recommended Decision offers only a vague and indefinite process for calculating rates- based support. It seems that the majority justifies the existing system on the ground that costs equal rates, and at the same time rejects all change on the ground that costs do not equal rates. In my view, so long as we continue to place primary reliance on cost- based support, the statute and the court decision require that we establish a dollar benchmark that is reasonably comparable to average urban costs. Similarly, the majority rejects the conclusion that an “urban cost benchmark would better satisfy the statutory comparison of urban and rural rates.” 187 Underlying 185 Recommended Decision, ¶ 41. 186 Recommended Decision, ¶ 39. 187 Recommended Decision, ¶ 39 (emphasis in original). 54 Federal Communications Commission FCC 02J- 2 10 this conclusion is the plausible assumption that explicitly relating the dollar benchmark to the national urban average cost would produce a lower dollar benchmark and an increase in support. The majority apparently finds that incremental support would be ineffective at producing comparable rates, but existing support passes the test. I fail to understand why this might be true. Once again, the majority reports that “proponents of the urban benchmark have not explained how additional funding produced by an urban benchmark would produce reasonably comparable rates.” 188 The explanation is obvious. Assuming that local rates are based on costs, support based on a comparable multiple of average urban costs will give each state an opportunity to set its own rates at an average level that is reasonably comparable to rates in urban areas of the nation. Other portions of the majority’s opinion precisely describe this relationship. 189 The majority criticizes proponents of an urban benchmark for not providing “a rational justification for setting the benchmark at any particular level.” 190 This imposes an unreasonable burden on those proponents, and is in fact a flaw in the majority’s own cost- based approach. As the majority notes (and the court explained) any determination of a benchmark “will necessarily be somewhat arbitrary.” 191 A choice obviously needs to be made from within a reasonable range of benchmarks. But this does not justify putting an unreasonable burden on the proponents of an urban benchmark. In my view, they have carried the entire burden that can fairly be assigned to them. They have shown convincingly that the existing benchmark does not provide sufficient support to make costs in high cost states reasonably comparable to national average urban costs. Moreover, the record does contain rational recommendations for a particular level. For example, three regulatory commissions from rural states recommended that the benchmark be set at a level no higher than 125 percent of average urban costs. 192 Even if the majority believes that this particular number is too low, that is no reason to leave in place a benchmark that the record shows is too high. 5. Evidence of Insufficiency 188 Recommended Decision, ¶ 40. 189 See, Recommended Decision, ¶ 25 (“ Despite implicit or explicit state support mechanisms, the low- cost areas of some states cannot balance their high cost areas. Although such states could, through their own efforts, achieve reasonably comparable rates within their own boundaries, those rates would still be high relative to the national average because of the states’ high average costs. The Commission’s primary role is to identify those states that do not have the resources within their borders to support all of their high- cost lines.”). 190 Recommended Decision, ¶ 40. 191 Qwest Corp. v. FCC, 258 F. 3d 1191, 1202 (10 th Cir. 2001); Recommended Decision, ¶ 33. 192 Comments of Maine Public Service Commission, Montana Public Service Commission, and Vermont Public Service Board at 7. (I do not participate in Montana proceedings concerning matters referred to the Universal Service Joint Board.) 55 Federal Communications Commission FCC 02J- 2 11 The preponderance of the credible evidence of record shows that current federal support, based on the 135 percent benchmark, does not provide support sufficient to attain reasonably comparable rates or costs between high- cost rural states and the nation’s urban areas. All of this evidence is based on the cost outputs from the Commission’s Synthesis Cost Model, but different methods have been used to select an “urban” sample. Despite the different samples, the results are remarkably consistent. Most notable was a filing from the Rural Utility Service (RUS), a branch of the United States Department of Agriculture. The RUS sampled 17 downtown urban areas to determine average urban cost. It determined that in those urban areas the average “loop cost” was $390 per line 193 and that the national average (urban and rural) “loop cost” was $672. Therefore a national average 135 percent benchmark amounts to 1.35 times $672 per line, or $907 per line. RUS reported this to be 233 percent of the urban “loop cost” of $390. RUS stated, and I agree, that this amount is far outside the range of reasonable comparability. Another approach would be to use the average cost among those companies that serve solely or primarily urban areas. Unfortunately, however, the value of this method is severely limited by the small sample available. Of 96 nonrural companies, only one even arguably meets the test of serving predominantly an urban area. It is C and P Telephone Company of Washington D. C., a Verizon affiliate. 194 The model currently shows that the average forward- looking cost of providing service in the District is $16.03. The current benchmark of $29.60 is 185 percent of this estimate of the national urban cost. Once again, this method shows that the current benchmark is far outside the range of reasonable comparability. Yet another model- based approach is to use the density zone feature of the cost model. 195 The FCC currently recognizes nine such zones. The three densest zones have a density of 2,550 lines per square mile or more. Together, these wire centers comprise 28 percent of all lines and in my view offer a reasonable set estimate of “urban” areas. 196 The weighted average cost in these three density zones is $16.34. The current benchmark of $29.60 is therefore 181 percent of national urban cost. Once again, this method shows 193 In most discussions of universal service, the standard unit of measure it total cost per line per month. The RUS used a different yardstick, but one still based upon the Commission’s own cost model. 194 Even C& P of D. C. may not be entirely “urban” since some census block groups in Northwest Washington are primarily suburban. 195 The Synthesis Model was originally designed with the capability of producing costs by density zone or by wire center, as the modeler may select. 196 At the lower edge of 2,550 lines per square mile, this amounts to approximately one- quarter acre per line. Since many residential and business customers have more than one line per occupied structure, this means that some areas in this zone have an average lot size larger than one- quarter acre. A generous definition of “urban” follows from using the higher costs associated with areas with large lot sizes. 56 Federal Communications Commission FCC 02J- 2 12 the benchmark is far outside the range of what might be considered reasonably comparable. Still another way to measure “urban” cost is to adopt the definitions of “urban” used by the United States Bureau of the Census. This analysis also confirms that the Commission is not today providing sufficient support. The Census Bureau defined urban areas for the 2000 census 197 as areas that are contained in either an “urbanized area” 198 or an “urban cluster.” 199 Staff from the Joint Board has utilized public domain data to identify those wire centers that are predominantly urban. 200 Ohio data were used because it happened that wire center boundaries for that state were available in the public domain. 201 The resulting estimate of urban average cost is $18.02. The current benchmark of $29.60 is therefore 164 percent of this estimate of national urban cost. Once again, this method shows that the benchmark is outside the range of reasonable comparability. In summary, using a variety of methods, the record reveals that the existing system supports net cost differences where rural areas have net costs that are somewhere between 164 percent of urban average cost and possibly as much as 233 percent of urban average cost. None of this evidence is consistent with a view or finding that the existing system is providing sufficient support to make net costs in high- cost states reasonably comparable to urban areas of the nation. For these reasons, I conclude that the existing 197 The Census Bureau noted that some agencies are required to that some Federal and state agencies are required by law to use Census Bureau- defined urban and rural classifications for allocating program funds, setting program standards, and implementing aspects of their programs. Federal Register: March 15, 2002, Volume 67, Number 51, Notices, pages 11663- 11670. 198 For Census 2000, an urbanized area consists of contiguous, densely settled census block groups (BGs) and census blocks that meet minimum population density requirements, along with adjacent densely settled census blocks that together encompass a population of at least 50,000 people. Id. Ohio has 19 urbanized areas. Internet cite: http:// www. census. gov/ geo/ www/ ua/ ua_ state_ corr. txt. 199 For Census 2000, an urban cluster consists of contiguous, densely settled census block groups and census blocks that meet minimum population density requirements, along with adjacent densely settled census blocks that together encompass a population of at least 2,500 people, but fewer than 50,000 people. Id. Ohio has 143 urban clusters. Internet cite: http:// www. census. gov/ geo/ www/ ua/ uc_ state_ corr. txt. 200 Predominantly urban was defined here as enclosing an area that is more than 50% urbanized area or urban cluster. 201 Because Ohio is a sizeable state with both rural and large city wire centers, it provides a good sample for urban costs nationally. Data for the entire country were available for $5,000, but the Joint Board declined to make this expenditure. For each of 903 wire centers in Ohio, staff determined the percentage of the wire center’s area that is characterized as “urbanized area” by the Census. 52 wire centers were then selected which serve areas that are at least 50 percent “urbanized areas.” Of these, 39 were served by nonrural carriers, and therefore average cost data were available. The average cost of these 39 wire centers, weighted for line size, was $17.63. 57 Federal Communications Commission FCC 02J- 2 13 system, based on a benchmark of 135 percent, does not meet the requirements of the statute. 6. Other Considerations Some may be reluctant to use an urban average or decrease the benchmark because they have limited confidence that some of the states receiving added support actually need that support. In other words, some may doubt that the states with high forward- looking cost actually need more support. Fortunately, several reasonable solutions are available to the Joint Board that should assuage that concern and allow the Joint Board to recommend a support mechanism that provides sufficient support to all. The first question is the accuracy of the model itself. One might support the underlying methodology of using forward- looking costs but might doubt the reliability of costs calculated under the existing cost model. In the nearly three years since issuance of the last significant order defining the model, 202 several potential problems with the model have been identified, 203 but the Joint Board has declined to investigate these matters. Opportunities to make significant improvements to the model have also been missed. 204 So long as the Commission continues to distribute support based on a forward- looking cost model, it should maintain a staff with expertise on the internal workings of the model, and it should devote substantial continuing resources toward improvement of that model. Moreover, the Joint Board should be actively involved in these activities to test, evaluate and improve the model. I am very encouraged that Joint Board chair Abernathy has committed to do this in the future. The broader question is the propriety of distributing support using costs that are different from the costs that states use to set rates. This concern is reasonable, because support and rates are calculated in most states using quite different costing rules. The Joint Board has previously recommended, and the Commission has adopted, a system of support for nonrural carriers that is based on forward- looking costs. As a result the 202 The last order establishing significant parameters of the cost model was the Tenth order, issued in November of 1999. 203 The Rural Task Force identified a number of areas in which the model appeared to produce unreasonable results for rural companies. In addition, there are potential problems with use of the model even for nonrural companies. For example, the model deploys loop plant without regard to natural and manmade barriers such as mountains, lakes and highways. Also, the model inputs now include many more special access lines than were originally anticipated, largely due to the unexpected sale of thousands of DS-1 and DS- 3 special access circuits. Some of these circuits are sold to competitors, others to schools and libraries using subsidies from federal universal service funds. We have not investigated whether those circuits are currently being assigned to the proper wire centers nor whether the model is properly adjusting switch, feeder and distribution plant to accommodate these new circuits. 204 More than one year ago FCC staff had developed a redesign of the model so that it constructed plant solely along rights of way. That modification significantly changed costs in some areas, but had the potential to improve the reliability of the model in areas with significant topographic limitations. However, all work stopped on those modifications when a key FCC staff member left the Commission. 58 Federal Communications Commission FCC 02J- 2 14 existing support system provides support to some areas with apparently low rates, and provides surprisingly little support to some areas with apparently high rates. Even if one holds this objection, I do not think it should stand in the way of providing more support here. One option would be to lower the benchmark but to deny support increases to carriers unless they also have above average embedded costs. Since all incumbent carriers still have an interstate revenue requirement, data on this question is already available through NECA. Thus embedded costs could be used as a screen to prevent forward- looking support from increasing increases in places that are likely to have low rates. More fundamentally, there may be general dissatisfaction with basing support on forward- looking costs because they have so little relation to customer rates. If so, this would justify a general review of the existing support mechanism. If we are unwilling or unable to conduct such a fundamental review, however, that is not a reason for us today to deny sufficient support to high- cost nonrural carriers. The majority has recommended no increase to cost- based funding for nonrural carriers. That analysis will comfort those who wish to ensure that the overall Universal Service fund does not grow as a result of this proceeding. To place this in context, however, it is important to understand the small effect this program has on the overall fund size. High cost fund support for nonrural carriers comprises only 4.3 percent of the total universal service support mechanisms. 205 Accordingly, even if forward- looking support to nonrural carriers were doubled (or if it were eliminated entirely) there would be only a small effect on the overall size of the fund. B. Rates- Based Support The court directed the Commission to explain how its cost- based support is designed to produce reasonably comparable rates. Instead, the Joint Board recommends adoption of an entirely new rates- based support mechanism characterized as a “supplemental rate comparability review”. 206 The rationale apparently is that even if the cost- based system does not provide sufficient support, then the availability of a supplemental proceeding can ensure compliance with the statute. Thus the Joint Board recommends that compliance with the statute no longer be measured solely by reference to a known objective formula against which results can be independently verified. Instead, it recommends today a new subjective system with rules that will be established on an ad hoc basis through agency case law and practice. For the reasons explained below, I conclude that creation of this new supplemental proceeding at this time does not comply with the Commission’s duties under section 254 and with the terms of the court’s remand directive. 1. Sufficiency of Cost- Based Support 205 FCC, Universal Service Monitoring Report, CC Docket No. 98- 202, Summer, 2002, Table 1.11. 206 Recommended Decision, ¶ 38. 59 Federal Communications Commission FCC 02J- 2 15 Section 254 became law six and one- half years ago. It contained language suggesting (but alas not requiring) that rates should promptly be made affordable, and rates and service reasonably comparable. For all those years the customers of high- cost nonrural carriers have lacked assurances that their rates and service are reasonably comparable to urban areas. The court has clearly directed the Commission to explain the relation between its cost- based support program and the statutes. Neither the Joint Board nor the Commission has ever found that current rates in all areas of the country are reasonably comparable to urban areas. Once again, the Joint Board fails to do so today. The Commission cannot avoid this statutory responsibility forever. Today the majority recommends essentially that the Commission go back and start again, basing sufficiency not on its well- defined (although deeply flawed) forward-looking cost- based system, but on a new and as yet undefined rates- based system. Having received a referral on this issue, I believe the Joint Board now has the obligation either to explain how cost- based support is sufficient or to change that support. The recommendation for a supplemental rates- based support system does not discharge that duty. 207 It is not an adequate response to tell customers, six and one- half years after the Act became law, after adopting a new cost- based support system, and after losing a remand from the Court of Appeals, that if customers or their service providers want sufficient support they may file a petition. I doubt such a plan would have been reasonable even in 1997 when the Commission first selected a broad direction for its support program. 208 To make this suggestion now, however, while making no change at all in the existing cost- based system, is an affront to those customers. One foreseeable effect will more years of delay before Section 254 is meaningfully implemented for customers of large carriers. I understand the majority to conclude that even if cost- based support is insufficient, section 254 can be satisfied if there exists a supplemental rates- based system that uses case- by- case determinations, and further that this new system need not publish in advance the rationales, algorithms or calculations that will be used to provide support. I am greatly concerned that this decision may be applied in the future to small rural carriers. While these carriers serve a minority of rural customers nationwide, in most states they serve areas that are predominantly rural. In many cases they have few or no low- cost customers upon whom to rely for low averaged rates. Accordingly, insufficient cost- based support is a problem for relatively few of the large nonrural carriers, but it can be a matter of great importance for the customers of small rural companies. Because I believe today’s Recommended Decision supports continuation of insufficient support, I 207 This principle may be of significance in the upcoming discussion of support for rural carriers. Rural carriers may be very interested in the establishment of a precedent that cost- based support need not, by itself, be sufficient to allow reasonably comparable rates. 208 I also note that the Commission did not provide support for nonrural carriers until 2000 because it felt it necessary to spend several years developing its new forward- looking cost model. 60 Federal Communications Commission FCC 02J- 2 16 believe it is a particularly worrisome development for rural carriers and their rural customers. 2. Burdening the States Under the majority’s opinion, states must apply for supplemental support. This unlawfully shifts the federal responsibility to the states, and it does not ensure that ratepayers in high cost states will receive sufficient support to achieve comparable urban and rural rates. Sufficient support will be provided only if the state chooses to seek it and if the state succeeds in making the required showings. A state might avoid such an undertaking for a variety of reasons. The proceedings will require considerable expertise and could be contentious. Many adjustments to rates are possible, but the parties are not given any guidance as to what adjustments will be accepted. Litigation costs, particularly for experts, could be sizeable. While some high cost states may have the resources and expertise to make the required showings, others are unlikely to accept this challenge. The majority may have misunderstood the court’s decision regarding inducements. I read the court’s decision to require the FCC to create inducements for the states to deal with any rate differences that may exist within their own borders. I do not believe that the court intended to permit the Commission to deny sufficient federal support to customers in a state where, for whatever reason, the state commission has elected not to apply for that support. 3. Appropriate State Roles The majority’s analysis repeatedly neglects the important difference between rate variation within a state and rate variation among the states. It is elementary that a dollar raised by a state universal service program is not only a dollar available for rate support in that state, but it is also a dollar that is added to local rates in that same state. Therefore, while state universal service funds can be highly effective at reallocating internal burdens within the state’s boundaries, they are powerless to reduce the state’s average rates or to change the state’s ranking among other states. Intrastate equity is within the sphere of what a state can and should accomplish. State- to- state equity is beyond any state’s means and has been recognized as the proper responsibility only of the Commission. Ominously, the blurring of this distinction occurs in the section on inducements. The Recommended Decision discusses the new supplemental proceeding in the section on inducements, thereby suggesting that there should be some sort of connection between a state’s performance of its own duties under section 254 and the availability of sufficient federal support. Indeed, the Recommended Decision says that an applicant should show it has “taken all actions reasonably possible and used all available state and federal resources to make basic service rates reasonably comparable, but that rates nevertheless 61 Federal Communications Commission FCC 02J- 2 17 fall above the benchmark.” 209 This might be a relatively harmless (although still facially improper) requirement if all it means is that the state must, one way or another, keep rates within its borders reasonably uniform. States have a variety of tools to accomplish this. 210 On the other hand, this language may be a warning that only states with substantial intrastate universal service fund mechanisms need apply. That would be a clear violation of section 254. A rural state having uniformly high rates and costs might need substantial federal support yet have no need for a state universal service fund because its rates are comparable within the state. Section 254 does not require states to establish universal service funds. 211 The Commission has no authority to make the existence of a state universal service fund a condition precedent to sufficient federal support. The “safe harbor” language also seems to be similarly confused about the extent of state duties. Apparently before a state may seek supplemental support it must first admit that it has anchored outside the newly announced “safe harbor.” Once again this overlooks the difference between intrastate and state- to- state rate variations. It may even prove to be a new kind of “Catch- 22” for applicants. To apply, the state must admit that its rates have fallen outside the “safe harbor.” This suggests that the Commission’s first inquiry in a supplemental support proceeding may be an inquisition into whether the state’s own universal service efforts have been adequate. 212 It remains to be seen how the Commission will respond to facts showing that the state has no internal high cost program because its rates are uniformly high. I believe today’s Recommended Decision should confirm earlier statements from this Joint Board and by the Commission that states are powerless to remedy state- to- state variations in rates and that this is the chief object of federal high cost policy. We should remove any suggestion that we believe that all states outside the “safe harbor” necessarily have some responsibility for that fact. Also, if there is going to be a supplemental proceeding, we should clarify that the proceeding has nothing to do with state inducements. 4. The New Proceeding The Recommended Decision provides few details about the new supplemental proceeding. The majority apparently envisions developing the details through 209 Recommended Decision, ¶ 50. 210 Most states rely on rate averaging for their nonrural carriers. Increasingly, however, states are replacing these systems with explicit universal service programs. 211 Section 254 provides that “A State may adopt regulations not inconsistent with the Commission's rules to preserve and advance universal service.” 47 U. S. C. § 254( f) (emphasis added). 212 Ironically, the requirement that rates fall outside the “safe harbor” may actually require states to deaverage rates in order to qualify for Federal USF. This would likely have the effect of reducing rate comparability in high cost areas, contrary to the statute. 62 Federal Communications Commission FCC 02J- 2 18 administrative case law and practice. 213 States must present a “rate analysis,” but they are also to be given “great flexibility” in their presentations. 214 Of primary importance here is the complexity of obtaining valid comparative rate measurements. As many as 30 different adjustments to raw rate levels may be necessary to make meaningful rate comparisons. Foreseeable adjustments may fall into five major groups: € Rate for Basic Local Service. Some customers face unavoidable extra charges for local exchange service. Other customers have simpler and more inclusive rate designs. Also, some customers have the option of choosing from a variety of local calling options. A particular customer may have a choice, for example, to select a more expensive plan that includes more local calls, more calling time or a wider calling area. € Quality of Service. Differences in service quality may be the result of different investment and maintenance decisions. € Calling Area Size. Some state calling area policies change the value of service to the customer in ways that do not appear in the nominal local exchange rate. The size of the local calling area is one important variable. Some customers have a large calling area that can extend for fifty miles or more. Customers in other areas, however, have relatively small calling areas, and they must pay toll rates when they make calls to local schools and nearby businesses, possibly even to Internet providers. For the latter group, toll bills are generally higher and local rates are generally lower. For this reason, two customers may have the same local exchange rates yet, because they have local calling areas of different sizes, comparisons of their local exchange services may not be valid. € Intrastate Toll and Access Rates. States differ in the degree to which they allocate joint and common costs to toll services. In some areas significant joint and common network costs are included in toll and access charges, and local exchange rates are low. In other states with different policies, local exchange rates would be higher, but toll rates would be lower. For this reason, two otherwise identical customers may have different local exchange rates based solely on state policy regarding the division of common costs between local and toll services. € State Regulatory Policies. Local exchange rates can in some instances reflect state regulatory policies that might be inappropriate to include in support calculations. A carrier may, for example, be earning an exceptionally high rate of 213 This can be contrasted with the “safety valve” mechanism adopted last year for rural carriers. That mechanism is clearly defined. Carriers are able to calculate exactly the amount of support available under that mechanism and to verify that the amount provided agrees with the Commission’s rule. 214 Recommended Decision, ¶ 56. 63 Federal Communications Commission FCC 02J- 2 19 return on its net plant investment. Another carrier may be using aggressive depreciation techniques for new equipment purchases. There are differences that may also be important related to capital structure, cost of debt, and interest synchronization. Given the stakes, states may, and undoubtedly will, present dozens of adjustments in their “rate analyses.” No two applicants will analyze their rates in the same way. One state may make large corrections for its relatively small calling areas. Another may emphasize one or another aspect of a rate design that permits customer to select from among several calling plans offering different kinds of prepaid services. 215 A third may argue that its high intrastate toll and access rates should be considered. The Commission already has experience with each state defining its own costs, if not rates; and it justifiably abandoned the experiment. In 1997 the Commission announced that states were encouraged to perform and file their own state- specific forward- looking cost studies, and that those studies might later be used to determine a state’s support. 216 Later, the Commission withdrew this offer, finding that only a uniform national- wide cost model could make appropriate support calculations. Individual state models, the Commission held: could rely on differing forward- looking cost methodologies, including differing assumptions or input data elements that would prevent meaningful comparisons of the resulting forward- looking cost estimates, and thus would provide a less accurate and consistent picture by which we could evaluate the cost levels that must be supported . . . . 217 The Joint Board here proposes a state- by- state approach with rates that is quite similar to the long- abandoned state- by- state cost studies. With each state free to suggest its own procedures for adjusting rates, it seems that there is a sizeable chance of uneven and even discriminatory results. Moreover, a state petitioner may not even have access to the data needed to prove its case. Suppose a state decides that it should adjust its local exchange rates to account for unusually small calling areas. Can it compare its own adjusted data with unadjusted rate data from other states? If not, how can it get adjusted data for other states? Today’s Recommended Decision offers no guidance about how these issues will be resolved. Applicants are not told whether they should adjust rates for local calling 215 For example, the GAO study noted that customers in Michigan had an alternative between one “message rate” calling plan costing $12.01 and another “unlimited basic service” plan costing $43.95. The GAO study used the higher figure. If Michigan were to apply for support, it would presumably use that same number. 216 First Order, ¶ 206; Second Order, ¶ 31. 217 Seventh Order, ¶ 52. 64 Federal Communications Commission FCC 02J- 2 20 area size, or how they should report rates when customers have choices among calling plans with different calling scopes or different numbers of included calls. 218 Who will win supplemental support? How will such support be calculated? The majority provides no answer. The Commission will have a very difficult time sorting through these competing claims and establishing the credibility of its new system as impartial and objective. Moreover, it will probably take at least another year before we find out how this new proceeding really will work. 5. Risks and Incentives A new and embryonic proceeding presents new risks and opportunities to the Commission and to the states. One risk is preferential or discriminatory treatment. Future Commissions have virtually unlimited discretion to define “rates” as they choose and then to award support or to withhold support, there is a substantial risk that they may consider factors not stated in a written decision. To avoid this kind of risk, government support programs of all sorts have been operated on a formula basis and have avoided case- based ad hoc decisions. 219 Apparently the majority does not consider preferential or discriminatory treatment a substantial risk. I mentioned above that the parties in a supplemental support proceeding may propose adjustments related to state regulatory policies such as authorized rates of return, depreciation, capital structure or interest synchronization. This will involve the Commission in reviewing the reasonableness of state rate decisions. If the rates actually being collected reflect a high “authorized rate of return” or high actual earnings, for example, the Commission may be asked to impute a lower rate of return. This would require it to substitute its judgment for that of the state commission. In the extreme, the Commission could find the proceeding grows essentially into an ersatz intrastate rate case. It may even be necessary to audit some company records. This would not only prove harmful to comity with the states, but it could be well beyond the actual staffing capacity of the Commission. A new rates- based support system could create perverse incentives for carriers. When it initially adopted a forward- looking cost model for support, the Commission expressed concern that an embedded- cost system would create undesirable incentives for incumbent carriers. If support were based on embedded cost, the Commission thought 218 Commendably, the Recommended Decision does suggest that there should be a time limit for such a proceeding. I agree that it is important that requests for supplemental support must be processed and decided expeditiously. Without such a limit, a proceeding could be initiated by a state or carrier, be noticed for comment, and then remain undecided for years. For example, the Vermont Public Service Board filed a petition for waiver of existing universal service rules in September of 1993. The Commission never acted on that petition. In re Waiver of Section 36.631 of the Commission’s Rules Governing the Universal Service Fund, Docket No. AAD 93- 103, public notice issued Oct. 15, 1993, 8 FCC Rcd. 7588. 219 Formula- based federal support is provided to the states across a wide range of public policy, including transportation support, farm supports, and health care support. 65 Federal Communications Commission FCC 02J- 2 21 that incumbent carriers would make inefficient investments and have a disincentive to efficient operations and new competitors would be deterred from entry. 220 But now the Joint Board proposes to base support upon the local rates that are, in most places, based on embedded costs. Once banished, the problem has now returned. The Commission may have to choose between initiating company audits or providing support for costs that are not essential to the services supported by universal service. Short of audits, nonrural companies may have practically unrestricted ability to generate support based on any infrastructure investments they desire and that are recognized by state commissions. A new rates- based support system also creates a new moral hazard by establishing perverse incentives for state commissions. Rates- based support will create an incentive for state commissions to maximize their support, even if it requires changes to local rate designs. Unless the method of measuring rates is carefully controlled, states may find themselves performing major redesigns of local rates. If for example, no adjustments are allowed for calling area size, states will have an incentive to increase calling areas. “One- state- one- rate” may become the motto of high- cost states around the country. Also, if no allowance is made for toll and access charges, states will have an incentive to reduce those charges and redesign their rates to increase local exchange charges. Finally the new system also presents a potentially issue to those who are concerned about the total size of the Universal Service Fund. Of course the entire support program for nonrural carriers amounts to less than five percent of the total universal service fund. Still, given the broad uncertainties in how local rates will be measured and how support will be calculated, the ultimate effect may be a significant increase in the high cost support for nonrural carriers. C. Transport Issue Finally, the transport issue has not been addressed here. In its supported services recommendation the majority promised to take up and address here the issue of high transport costs. Without support, those companies with high transport cost will not receive sufficient support to allow their rates to be comparable with urban areas. We recently issued a Recommended Decision on supported services. One of the questions discussed there was federal support for transport costs. I said at the time that I wanted to examine this issue as soon as possible because very high unsupported transport costs are the primary barrier to establishing telephone service in certain remote areas. The majority ultimately stated that we would analyze the question of whether support should be modified for transport costs, and that the analysis would occur here, in the context of the Tenth Circuit remand. 221 We have not performed that analysis, and I regret 220 Federal- State Joint Board on Universal Service, CC Docket No. 96- 45, Report and Order, 12 FCC Rcd 8776 (1997) (Universal Service Order), as corrected by Federal- State Joint Board on Universal Service, Errata, CC Docket No. 96- 45, FCC 97- 157 (rel. June 4, 1997) (hereinafter “First Report and Order”) ¶ 228. 221 In that Recommendation the majority said: (continued….) 66 Federal Communications Commission FCC 02J- 2 22 the omission. I continue to believe that high transport costs must be examined immediately to ensure compliance with section 254 of the Act. 222 (Continued from previous page) The Joint Board recognizes that issues related to the cost of transport facilities are important and may be of particular concern in rural areas of rural and non- rural companies. Transport may be a necessary element of a carriers’ provision of services eligible for federal universal service support. In some remote communities, customers must use the interexchange network to access essential community services such as law enforcement, health care, schools and libraries. The extent to which these costs should be supported is best addressed in our recommendation on the decision remanded from the Tenth Circuit where we will analyze reasonable comparability and sufficiency of support. In that proceeding, we will consider the level of support necessary to ensure that all citizens of the United States have access to reasonably comparable communications services. Recommended Decision FCC 02J- 1, ¶ 57 (emphasis added). 222 See, Separate Statement of Bob Rowe, Part VI. 67