Docstoc

ohio

Document Sample
ohio Powered By Docstoc
					              BEFORE THE PUBLIC UTILITIES COMMISSION OF OHIO




In the Matter of the Review Of SBC Ohio’s TELRIC )
Costs for Unbundled Network Elements             )   Case No. 02-1280-TP-UNC




               DIRECT TESTIMONY OF TREVOR R. ROYCROFT, Ph.D.


                                     ON BEHALF OF


                   OFFICE OF THE OHIO CONSUMERS’ COUNSEL
                          10 West Broad Street, Suite 1800
                             Columbus, Ohio 43215-3485
                                  (614) 466-8574




                                   PUBLIC VERSION




Dated: May 28, 2004
Table of Contents

Statement of Qualifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Tab 1: Overview and Summary of Testimony . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
        Policy Framework: UNEs, Monopoly, and Economic Cost . . . . . . . . . . . . . . . . . . . . . . . . . 11

Tab 2: Dr. Aron, UNE Rates, Rate-of-Return Regulation and Costs . . . . . . . . . . . . . . . . . . . . . . . . 20
       The Desirability of Facilities-Based Competition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
       Dr. Aron and Fill Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
       Dr. Aron and Depreciation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

Tab 3: Evaluation of SBC Ohio’s Cost Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
       Cable Sizing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
       The FCC’s Perspectives on Fill . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
       Fill Factors and Operating Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
       TSLRIC and TELRIC Fills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
       Other Issues with LoopCAT and Distribution Plant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
       Improvements in LoopCAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
       Alternative Cost Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
       Unifying Assumptions Across the Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

Tab 4: Shared and Common Costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

Tab 5: Dr. Vanston and Economic Depreciation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
       Wireless Substitution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
       Cable Companies and Voice Over IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
       The Impact of Competition on the Economic Value of SBC Facilities . . . . . . . . . . . . . . . . 116
       Technological Change and Economic Value—The Impact of Broadband . . . . . . . . . . . . . . 118
       The FCC’s Current Depreciation Lives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
       SBC Ohio’s Proposed Depreciation Lives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125




                                                                      i
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


                    DIRECT TESTIMONY OF TREVOR R. ROYCROFT, Ph.D.
                                    ON BEHALF OF
                         OFFICE OF OHIO CONSUMERS’ COUNSEL

                                       CASE NO. 02-1280-TP-UNC

                                               MAY 28, 2004



     Statement of Qualifications

 1   Q1:   PLEASE STATE YOUR NAME AND ADDRESS.

 2   A1:   My name is Trevor R. Roycroft and my address is 9065 Echo Lane, Athens, Ohio, 45701.




 3   Q2:   WHAT IS YOUR OCCUPATION?

 4   A2:   I am an Associate Professor in the J. Warren McClure School of Communication Systems

 5         Management at Ohio University.




 6   Q3:   PLEASE DESCRIBE YOUR EDUCATIONAL BACKGROUND.

 7   A3:   In June 1984 I received the Bachelor of Arts degree in Economics with a minor in Statistics from

 8         California State University, Sacramento. The degree was awarded with honors. In September

 9         of 1986 I received the Master of Arts degree in Economics from the University of California,

10         Davis. In December of 1989 I received the Doctor of Philosophy in Economics from the

11         University of California, Davis. My Ph.D. fields of specialization are Economic Theory, Industrial

12         Organization, Public Sector Economics, and Economic History.




13   Q4:   PLEASE DESCRIBE YOUR PROFESSIONAL BACKGROUND.

14   A4:   I have been employed by Ohio University in the school of Communication Systems Management

                                                       1
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


 1         since the Fall of 1994. I was granted tenure and promoted to the Associate Professor rank in the

 2         Spring of 2000. My primary areas of teaching responsibility are graduate and undergraduate

 3         courses covering regulatory policy, the economics of the telecommunications industry, and

 4         telecommunications technology. Prior to my employment at Ohio University, I was employed by

 5         the Indiana Office of Utility Consumer Counselor (OUCC) from May of 1991 until July of 1994.

 6         For most of my tenure at the OUCC I was Chief Economist. Prior to my employment at the

 7         Indiana OUCC, I was a Visiting Assistant Professor of Economics at Kenyon College, Gambier,

 8         Ohio, during the 1989 and 1990 academic years. I have also been an Instructor of Economics at

 9         California State University, Sacramento in the Fall of 1987 and during the 1988 academic year.

10         Economics courses that I have conducted include: Principles of Economics, Intermediate

11         Microeconomics, Industrial Organization, Economics of the Public Sector, and Economics of

12         Developing Countries. Attachment TRR-1 is a copy of my vita.




13   Q5:   HAVE YOU PREVIOUSLY PRESENTED TESTIMONY BEFORE THIS COMMISSION?

14   A5:   Yes, I have. A complete listing of testimony is provided in my vita, which is provided as

15         Attachment TRR-1.




16   Q6:   WHAT IS THE PURPOSE OF YOUR TESTIMONY IN THIS CASE?

17   A6:   The purpose of this testimony is to evaluate SBC Ohio’s cost studies supporting its request for an

18         increase in UNE prices. At this phase of the proceeding the matters before the Commission relate

19         to prices for UNE loops. The primary focus of my testimony will be to evaluate arguments offered

20         by SBC Ohio witnesses Dr. Aron, Mr. Smallwood, Mr. Hamrock, Dr. Vanston, and Dr. Currie



                                                       2
                                                                   Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                              On Behalf of OCC
                                                                                   Case No. 02-1280-TP-UNC
                                                                                                 Public Version


 1         with regard to the rate increase requested by SBC Ohio. While there are a variety of loop types

 2         under consideration in this proceeding, the focus of my analysis will be directed at the 2-wire

 3         analog loop study.




 4   Q7:   WHAT HAVE YOU DONE TO PREPARE FOR YOUR TESTIMONY?

 5   A7:   I have reviewed SBC Ohio’s filing in this Case. I have prepared discovery requests which were

 6         served on SBC Ohio. I have reviewed the responses to those requests and have prepared follow-

 7         up discovery as appropriate. I have reviewed discovery requests served by other parties to this

 8         proceeding, and have reviewed relevant responses to that discovery. I have reviewed the

 9         deposition transcripts of SBC Ohio witnesses, and I attended the deposition of James Smallwood.

10         I have reviewed SBC Ohio’s LoopCAT model and the documentation provided with that model.

11         I have conducted multiple runs of the LoopCAT model which I used to assess the sensitivity of the

12         model’s outputs to changes in various assumptions. I have used the FCC’s Synthesis Model, run

13         with SBC-Ohio-specific input values to benchmark the performance of the LoopCAT model. I

14         have reviewed the Commission’s TELRIC-related rules. 1 I have reviewed the FCC’s Local

                                                                                                            2
15         Competition Rules, the FCC’s Triennial Review Order, and the FCC’s TELRIC NPRM.                      I


           1
                   In the Matter of the Commission Ordered Investigation of the Existing Local
                   Exchange Competition Guidelines, Case No. 99-998-TP-COI. In the Matter of
                   the Commission Review of the Regulatory Framework for Competitive
                   Telecommunications Services Under Chapter 4927, Revised Code, Case No 99-
                   563-TP-COI.
           2
                   In the Matter of Implementation of the Local Competition Provisions of the
                   Telecommunications Act of 1996; Interconnection between Local Exchange
                   Carriers and Commercial Mobile Radio Service Providers. CC Dockets 96-98
                   and 95-185. FCC Order No. 96-325. August 8, 1996. ¶29. (Hereinafter, Local
                   Competition Rules.) In the Matter of Review of the Section 251 Unbundling
                                                                                       (continued...)

                                                      3
                                                                   Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                              On Behalf of OCC
                                                                                   Case No. 02-1280-TP-UNC
                                                                                                 Public Version


 1         have reviewed the U.S. Supreme Court’s rulings in the Iowa Utilities Board and Verizon cases.3




 2   Q8:   HOW IS YOUR TESTIMONY ORGANIZED?

 3   A8:   I have organized my testimony into five sections which are separated by tabs. Tab 1 provides the

 4         summary of my findings and recommendations and provides a general overview of the major policy

 5         issues in this proceeding. Tab 2 addresses policy arguments offered by Dr. Aron which indicate

 6         that UNE rates should be set to “compensatory” levels. Tab 3 presents the results of my analysis

 7         of SBC Ohio’s cost model, the LoopCAT model. In this section I compare the results from

 8         LoopCAT with those of the FCC Synthesis model, a model which is known to be consistent with

 9         economic cost principles. Tab 4 examines SBC Ohio’s shared and common cost recovery

10         proposal. In this section I will identify problems with SBC Ohio’s approach that result in an

11         inappropriate overstatement of shared and common costs.            Tab 5 addresses economic

12         depreciation within a forward-looking cost model. This section addresses the testimony of Mr.

13         Vanston and analyzes two areas which have the potential to impact economic value of SBC Ohio


           2
            (...continued)
                    Obligation of Incumbent Local Exchange Carriers, Implementation of the Local
                    Competition Provisions of the Telecommunications Act of 1996, Deployment of
                    Wireline Services Offering Advanced Telecommunications Capability. CC
                    Dockets Nos. 01-338, 96-98, 98-147. August 21, 2003. (Hereinafter, Triennial
                    Review Order.) In the Matter of Review of the Commission’s Rules Regarding
                    the Pricing of Unbundled Network Elements and the Resale of Service by
                    Incumbent Local Exchange Carriers, WC Docket No. 03-173. FCC 03-224,
                    September 15, 2003. (Hereinafter, TELRIC NPRM.)
           3
                   IOWA UTILITIES BOARD, ET AL. V. FEDERAL COMMUNICATIONS
                   COMMISSION, ET AL. 1999 WL 24568 (U.S.) (Hereinafter, Iowa Utilities
                   Board).
                   VERIZON COMMUNICATIONS INC. V. FCC (00-511)219 F.3d 744, affirmed in
                   part, reversed in part, and remanded. May 13, 2002, pp. 40-41. (Hereinafter,
                   Verizon Ruling.)

                                                      4
                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                  On Behalf of OCC
                                                                       Case No. 02-1280-TP-UNC
                                                                                     Public Version


1   facilities—competition and technological change.




                                              5
                                    TAB 1

Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


 1   Tab 1: Overview and Summary of Testimony

 2   Q9:   PLEASE SUMMARIZE YOUR TESTIMONY.

 3   A9:   My testimony begins with an analysis of the appropriate standard to apply when determining UNE

 4         rates for SBC Ohio. My analysis leads me to the following conclusions on this matter.

 5         •      The approach to establishing prices for UNEs like 2-wire analog loops should be firmly
 6                rooted in forward-looking economic cost methodology. The FCC’s Total Element Long
 7                Run Incremental Cost (TELRIC) approach and the Ohio Commission’s cost guidelines
 8                identify an appropriate general approach for determining UNE prices.

 9         •      Recent pronouncements from the FCC regarding the potential to modify its TELRIC rules
10                should have no bearing on the conclusions reached in this proceeding. It is highly
11                speculative at this time to anticipate what changes in TELRIC the FCC might implement,
12                and one must also recognize that any decision made by the FCC will be subject to judicial
13                review.

14         •      Even if there are changes made in TELRIC as a result of the FCC’s review of the
15                methodology, it is highly likely that notions such as the optimization of network design and
16                a general forward-looking approach to developing UNE prices will continue to be applied.
17                The Telecommunications Act of 1996 explicitly prohibits application of rate-of-return
18                regulation, thus absent the development of some wholly different non-rate-of-return
19                approach, economic cost methods are still likely to be relied upon.

20                With regard to my analysis of SBC Ohio’s filing, I have reached the following conclusions:

21         •      SBC Ohio’s justification for requesting the substantial UNE rate increases is based on an
22                analysis of historical costs and a demand for UNE prices which will make SBC Ohio
23                shareholders whole. Dr. Aron’s claim that current UNE rates are not “compensatory” is
24                based on an analysis of SBC’s Ohio’s historical costs. Because there is not a predictable
25                relationship between economic costs and historical costs, Dr. Aron’s attempt to “validate”
26                the outcome of an economic cost model by way of historical cost data is incorrect and
27                misleading.

28         •      Dr. Vanston’s analysis of market and technological forces does not reflect a reasonable
29                interpretation of facts related to the impact of these forces on the economic value of SBC
30                Ohio’s facilities. Dr. Vanston’s analysis is highly speculative and relies on conjecture
31                instead of analysis of actual market conditions.

32         SBC Ohio’s cost model for estimating the direct costs of UNE loops, the LoopCAT model, as

33         filed by the Company, is biased toward the recovery of embedded costs and thus does not provide

                                                      6
                                                             Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                        On Behalf of OCC
                                                                             Case No. 02-1280-TP-UNC
                                                                                           Public Version


 1   the appropriate means to develop forward-looking economic costs. Problems with the LoopCAT

 2   model are associated with the user-adjustable inputs placed in the model by SBC Ohio, and with

 3   the structure of the model.


 4   Some of the problems with the user-adjustable inputs utilized in LoopCAT include:

 5   •       Fill factors which are based on the embedded base of SBC Ohio’s outside plant facilities.
 6           These fill factors are not forward-looking and are thus inconsistent with estimating
 7           economic costs which can then serve as a foundation for TELRIC prices.

 8   •       Depreciation rates which are based on financial reporting lives. Financial reporting lives
 9           are based on principles of financial conservatism, not methods associated with estimating
10           forward-looking economic costs. As a result, cost estimates based on these depreciation
11           rates are inconsistent with estimating economic costs which can then serve as a foundation
12           for TELRIC prices. SBC Ohio justifies the shorter depreciation lives used in its cost
13           modeling on a highly speculative and unrealistic assessment of the progress of local
14           exchange competition and on a highly speculative and unrealistic assessment of the
15           potential for technological substitution for SBC Ohio’s facilities.

16   •       Capital costs which are not forward looking are utilized to establish the annual charge
17           factors needed to develop recurring costs. As a result, cost estimates based on these
18           capital costs are inconsistent with estimating economic costs which can then serve as a
19           foundation for TELRIC prices.

20   In addition to the problems with LoopCAT which are associated with the user-adjustable inputs

21   discussed above, I have uncovered other structural problems with the model. Some of the

22   problems with the structure of the LoopCAT model include:

23   •       LoopCAT’s use of embedded investment information, which prevents the appropriate
24           forward-looking orientation of the cost model.

25   •       LoopCAT’s inability to optimize cable sizes.

26   •       LoopCAT’s inability to report results on a wire center basis.

27   •       LoopCAT’s inappropriate method of estimating the length of distribution cable, an
28           approach which is not based on information specific to SBC Ohio’s network area and
29           which leads to an overstatement of distribution cable length and investment, and thus
30           increases cost estimates.

                                                 7
                                                               Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                          On Behalf of OCC
                                                                               Case No. 02-1280-TP-UNC
                                                                                             Public Version


 1   •       An inappropriate incorporation of the impact of Multiple Dwelling Units (MDUs) into the
 2           cost modeling process. SBC Ohio utilizes statewide averages to estimate the impact of
 3           MDUs on costs. This approach ignores the higher distribution of multiple dwelling units
 4           in urban areas and is thus likely to result in higher UNE loop costs in higher density zones.

 5   •       An inappropriate inclusion of the costs of Feeder Distribution Interfaces (FDIs) for loops
 6           which terminate without the use of distribution plant. This will result in increased loop
 7           costs.

 8   I have also found problems with SBC Ohio’s approach to modeling shared and common costs.

 9   SBC Ohio’s basic approach relies on the creation of a “Common Cost Factor” and “Shared Cost

10   Factor” based on separate ratios of total common costs to total direct costs, and total shared costs

11   to total wholesale direct costs. SBC Ohio uses these ratios to gross-up direct cost estimates. With

12   regard to the creation and application of these Cost Factors I have found:


13   •       SBC Ohio has inappropriately relied on historical costs to develop the numerators in both
14           the Common Cost Factor and Shared Cost Factor.

15   •       SBC Ohio has developed the denominators in the Common Cost Factor and Shared Cost
16           Factor by adjusting historical cost information to approximate “forward-looking” costs.

17   •       The mixing and matching of the historical-cost numerator and a “forward-looking” cost
18           denominator is not appropriate. The ratios used to generate the Cost Factors should be
19           based on forward-looking cost estimates. Absent the ability to generate a reliable
20           forward-looking cost estimate, appropriate adjustments should be made to historical cost
21           levels to approximate a forward-looking view.

22   •       SBC Ohio’s mix and match approach to creating the Cost Factors increases the overall
23           Shared and Common Cost factor by about 4.5 percentage points (representing an increase
24           of approximately 22% for the overall factor) as compared to a pure historical cost
25           approach. Even SBC Ohio agrees that forward-looking shared and common costs are
26           lower than historical, thus it is likely that forward-looking Shared and Common Cost
27           Factors would be lower still.

28   •       Great care should be made in the application of any Shared and Common Cost Factor to
29           the unadjusted output of the LoopCAT model. Without adjustment, the LoopCAT model
30           produces cost estimates which include costs that are not forward looking and which reflect
31           embedded investment and expense levels. Thus, even properly constructed Shared and
32           Common Cost Factors applied to unadjusted LoopCAT results will result in excessive

                                                  8
                                                               Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                          On Behalf of OCC
                                                                               Case No. 02-1280-TP-UNC
                                                                                             Public Version


 1             recovery of embedded costs in UNE loop rates.




 2   I have conducted a benchmark analysis of the LoopCAT model using the FCC’s Synthesis Model

 3   (FCC SM). The FCC SM is a cost model which was developed by the FCC to assist it in

 4   establishing universal service funding levels for large LECs, including the RBOCs. Because it is a

 5   model which generates forward-looking economic costs, it can also be applied to generate cost

 6   estimates associated with UNEs. Recently the FCC Staff applied the FCC SM to settle a dispute

 7   between AT&T and WorldCom and Verizon regarding UNE pricing in the Virginia Arbitration

 8   Order.4




 9   In preparing this analysis I unified a large number of input values across the LoopCAT model and

10   FCC SM. In addition to unifying the “big three” assumptions of fill factors, depreciation rates, and

11   capital costs, I also updated the FCC SM’s user adjustable inputs with proprietary prices drawn

12   from the LoopCAT model for a variety of major network components. This analysis has allowed

13   me to assess the LoopCAT model in the light of a cost model which is known to be consistent with

14   forward-looking economic costing principles. My analysis leads me to conclude:

15   •         That the bias of the LoopCAT model toward embedded costs results in higher cost



     4
               In the Matter of Petition of WorldCom, Inc. Pursuant to Section 252(e)(5) of the
               Communications Act for Preemption of the Jurisdiction of the Virginia State
               Corporation Commission Regarding Interconnection Disputes with Verizon
               Virginia Inc., and for Expedited Arbitration, CC Docket No. 00-218, and In the
               Matter of Petition of AT&T Communications of Virginia Inc., Pursuant to Section
               252(e)(5) of the Communications Act for Preemption of the Jurisdiction of the
               Virginia State Corporation Commission Regarding Interconnection Disputes with
               Verizon Virginia Inc. CC Docket No. 00-251, DA 03-2738, Memorandum Opinion
               and Order, August 29, 2003. (Hereinafter, Virginia Arbitration Order.)

                                                  9
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1                  estimates than are observed to be the case for a model which is known to be consistent
 2                  with forward-looking cost principles.

 3          •       The LoopCAT exhibits a bias which leads to relatively higher cost estimates for higher
 4                  density zones. I believe that this outcome is the result of LoopCAT’s inappropriate
 5                  treatment of FDIs and MDUs and is also influenced by the LoopCAT’s defective method
 6                  of estimating distribution cable length.

 7   Q10: GIVEN YOUR ANALYSIS, WHAT DO YOU RECOMMEND?

 8   A10:   Based on my evaluation of the SBC Ohio filing, SBC Ohio has not justified the UNE rate increases

 9          that it has requested. Given that SBC Ohio’s case is based on the misplaced application of

10          arguments and techniques which are associated with embedded cost and rate-of-return regulation,

11          the Commission should reject SBC Ohio’s claims.




12          I have established a reasonable range of forward-looking cost values for 2-wire analog loops. The

13          lower-end of this range is based on cost estimates produced by the FCC SM. In addition, I

14          believe that modifications can be made to SBC Ohio’s LoopCAT which will allow it to develop

15          UNE rates which are more consistent with TELRIC. I use the modified LoopCAT cost estimates

16          to establish the upper-end of the reasonable range of direct costs for 2-wire analog loops. The

17          adjustments I made to LoopCAT include:

18          •       Use of the cost of capital estimate developed by OCC Witness Dr. J. Randall Woolridge.

19          •       The use of depreciation lives and salvage values which are consistent with those prescribed
20                  by the FCC. The values that I have used in my modeling, as shown in Attachment TRR-2,
21                  are appropriate for generating forward-looking economic costs.

22          •       The use of fill factors which are not based on SBC Ohio’s historic network investments.
23                  The fill factors which I used in my alternative LoopCAT run, as shown in Attachment
24                  TRR-2, are appropriate.

25          In addition, I made other adjustments to the LoopCAT results. I used my benchmark analysis to



                                                       10
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          true-up LoopCAT’s direct cost estimates to be consistent with the cost structure of UNE rates

 2          identified by the FCC SM.




 3          Finally, with regard to shared and common costs, I recommend the following:

 4          •       The creation of shared and common cost factors which reflect a consistent application of
 5                  either forward-looking costs, or a reasonably-adjusted application of historic costs.


 6   Policy Framework: UNEs, Monopoly, and Economic Cost

 7   Q11: WHAT IS THE BASIC ISSUE RELATING TO THE NEED TO ESTABLISH PRICES FOR
 8        SBC OHIO’S UNBUNDLED NETWORK ELEMENTS (UNEs)?

 9   A11:   The basic issue is the continuing existence of SBC Ohio’s monopoly power with regard to the

10          provision of UNEs. While the Telecommunications Act of 1996 mandates the provision of UNEs

11          by ILECs such as SBC Ohio, the courts have interpreted this requirement to only be associated

12          with UNEs where continuing barriers to alternative sourcing or self-provision exist.

13                  Section 251(d)(2) requires the FCC to determine on a rational basis which network
14                  elements must be made available, taking into account the 1996 Act’s objectives and giving
15                  some substance to the “necessary” and “impair” requirements.5

16          If UNEs can be obtained from other sources, or economically self-provided by CLECs, then the

17          need for unbundling is eliminated. Where it can be shown that the “necessary and impair” threshold

18          has been met for a particular UNE, then regulation of the UNE’s rates terms and conditions

19          requires the exercise of regulatory oversight to prevent the abuse of the continuing monopoly power

20          in the provision of the UNEs in question.




            5
                    Iowa Utilities Board, et al. v. FCC. §2(b).

                                                        11
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q12: ACCORDING TO THE TELECOMMUNICATIONS ACT, HOW IS THE REGULATORY
 2        OVERSIGHT TO BE EXERCISED WHERE MONOPOLY PROVISION OF UNEs
 3        CONTINUES TO BE AN ISSUE?

 4   A12:   The Telecommunications Act identified a set of criteria which was offered as guidance to the FCC,

 5          the agency ultimately charged with developing rules to implement the provisions of the Act. With

 6          regard to UNEs, the Telecommunications Act specifies:

 7                  PRICING STANDARDS-
 8                  (1)      INTERCONNECTION                     AND        NETWORK             ELEMENT
 9                  CHARGES—Determinations by a State commission of the just and reasonable rate for
10                  the interconnection of facilities and equipment for purposes of subsection (c)(2) of section
11                  251, and the just and reasonable rate for network elements for purposes of subsection
12                  (c)(3) of such section--
13                  (A) shall be--
14                           (i) based on the cost (determined without reference to a rate-of-return or other
15                           rate-based proceeding) of providing the interconnection or network element
16                           (whichever is applicable), and
17                           (ii) nondiscriminatory, and
18                  (B) may include a reasonable profit.

19          By excluding rate-of-return (ROR) regulation from the rate setting process, the Act thus turned the

20          attention of regulatory agencies to economic cost estimation. The FCC’s Total Element Long Run

21          Incremental Cost (TELRIC) appropriately rejected ROR approaches and incorporated the

22          appropriate forward-looking economic cost perspective:

23                  While we are adopting a version of the methodology commonly referred to as TSLRIC as
24                  the basis for pricing interconnection and unbundled elements, we are coining the term “total
25                  element long run incremental cost” (TELRIC) to describe our version of this methodology.
26                  The incumbent LEC offerings to be priced using this methodology generally will be
27                  “network elements,” rather than “telecommunications services,” as defined by the 1996
28                  Act. More fundamentally, we believe that TELRIC-based pricing of discrete network
29                  elements or facilities, such as local loops and switching, is likely to be much more
30                  economically rational than TSLRIC-based pricing of conventional services, such as
31                  interstate access service and local residential or business exchange service.6




            6
                    Local Competition Order, ¶678.

                                                        12
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          The U.S. Supreme Court was asked to render an opinion on the FCC’s interpretation of the

 2          pricing provisions of the Act.        The Supreme Court recognized the objectives of the

 3          Telecommunications Act with regard to price control stretched beyond those associated with

 4          typical regulatory objectives surrounding “just and reasonable” rates:

 5                  For the first time, Congress passed a ratesetting statute with the aim not just to balance
 6                  interests between sellers and buyers, but to reorganize markets by rendering regulated
 7                  utilities’ monopolies vulnerable to interlopers, even if that meant swallowing the traditional
 8                  federal reluctance to intrude into local telephone markets. The approach was deliberate,
 9                  through a hybrid jurisdictional scheme with the FCC setting a basic, default methodology
10                  for use in setting rates when carriers fail to agree, but leaving it to state utility commissions
11                  to set the actual rates.

12                  While the Act is like its predecessors in tying the methodology to the objectives of “just
13                  and reasonable” and nondiscriminatory rates. . . . It is radically unlike all previous statutes
14                  in providing that rates be set “without reference to a rate-of-return or other rate-based
15                  proceeding”. . . . The Act thus appears to be an explicit disavowal of the familiar public-
16                  utility model of rate regulation (whether in its fair-value or cost-of-service incarnation)
17                  presumably still being applied by many States for retail sales. . .7

18          Thus, under the Telecommunications Act prices for UNEs must be established without reference

19          to rate-of-return regulation. This divorce from familiar rate-setting methods, however, places a firm

20          like SBC Ohio under differing regulatory structures for its retail and wholesale operations. Retail

21          operations governed by a price cap plan may result in prices which diverge from economic cost.

22          Wholesale prices for UNEs, however, must be built on economic costs. In either case, rate-of-

23          return regulation is absent as a mechanism governing prices.




24   Q13: WHAT IS THE REASON FOR USING ECONOMIC COST METHODOLOGY?

25   A13:   Given the lack of competitive alternatives for UNEs regulators must establish price levels, because



            7
                    Verizon Communications, et al. v. Federal Communications Commission, et al.,
                    pp. 16-17.

                                                          13
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          market forces are incapable of doing so. If the market for UNEs were in fact competitive, it would

 2          be expected that in the long run, UNE prices would be driven toward long-run incremental cost.

 3          In other words, if the market for UNEs was competitive, prices would behave as is observed in

 4          other markets where competition is present. Thus, the use of economic cost methodology, such

 5          as TELRIC, will enable regulators to develop prices which will be reasonably consistent with what

 6          would be expected in a competitive market. Applying techniques associated with rate-of-return

 7          regulation would not result in an outcome which would be consistent with that expected in a

 8          competitive market. Thus, the rate-of-return approach, while being explicitly prohibited under the

 9          Telecommunications Act, is also inferior from the standpoint of the likelihood that an economically

10          inefficient outcome would be forced on the market.


11   Q14: IS THERE A “BRIGHT LINE” BETWEEN ECONOMIC COST METHODS LIKE TELRIC
12        AND HISTORICAL COST METHODS ASSOCIATED WITH RATE-OF-RETURN
13        REGULATION?

14   A14:   From a theoretical perspective, yes. However, when moving from theory to practice the

15          identification of a “bright line” is made more complex due to the fact that calculating economic costs

16          utilizes cost models that employ a large number of assumptions. While it might be the case that any

17          one of the assumptions in isolation is not capable of providing some sort of “bright line” standard,

18          as a general proposition, the more a cost estimate relies on information which reflects embedded

19          network investments and operations, the less likely the resulting cost estimate is to reflect forward-

20          looking economic costs. This point can be illustrated by considering the FCC’s logic when it

21          developed the TELRIC approach. The FCC identified three possible approaches to the

22          calculation of forward-looking economic costs.

23          (1)     Under the first approach, the forward-looking economic cost for interconnection and


                                                         14
                                                               Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                          On Behalf of OCC
                                                                               Case No. 02-1280-TP-UNC
                                                                                             Public Version


 1           unbundled elements would be based on the most efficient network architecture, sizing,
 2           technology, and operating decisions that are operationally feasible and currently available
 3           to the industry.8

 4   (2)     Under the second approach, the costs of interconnection and unbundled network elements
 5           would be based on existing network design and technology that are currently in operation.
 6           Because this approach is not based on a hypothetical network in the short run, incumbent
 7           LECs could recover costs based on their existing operations, and prices for
 8           interconnection and unbundled elements that reflect inefficient or obsolete network design
 9           and technology. This is essentially an embedded cost methodology.9

10   (3)     Under the third approach, prices for interconnection and access to unbundled elements
11           would be developed from a forward-looking economic cost methodology based on the
12           most efficient technology deployed in the incumbent LEC’s current wire center locations.
13           This approach mitigates incumbent LECs’ concerns that a forward-looking pricing
14           methodology ignores existing network design, while basing prices on efficient, new
15           technology that is compatible with the existing infrastructure. This benchmark of forward-
16           looking cost and existing network design most closely represents the incremental costs that
17           incumbents actually expect to incur in making network elements available to new entrants.
18           . . . We, therefore, conclude that the forward-looking pricing methodology for
19           interconnection and unbundled network elements should be based on costs that
20           assume that wire centers will be placed at the incumbent LEC’s current wire center
21           locations, but that the reconstructed local network will employ the most efficient
22           technology for reasonably foreseeable capacity requirements.10


23   Thus, while the FCC’s approach to the development of economic costs was not as pure as it might

24   have been, it generally avoids the use of embedded costs. It is important to note from ¶685 of the

25   FCC’s Local Competition Order, cited immediately above, that the incremental costs which the

26   incumbents “actually expect to incur” are those associated with a reconstructed local network that

27   will employ “the most efficient technology for reasonably foreseeable capacity requirements,” not

28   the actual costs of their existing network. The notion of “actual costs” is one that is raised by SBC



     8
             Local Competition Order, ¶683.
     9
             Local Competition Order, ¶684.
     10
             Local Competition Order, ¶685, emphasis added.

                                                 15
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          Ohio witnesses and actual costs must be put into their appropriate context, i.e., the FCC’s

 2          statement regarding the costs that incumbents “actually expect to incur” rejects the embedded cost

 3          approach. SBC Ohio witnesses inappropriately associate actual costs with embedded costs and

 4          rate-of-return regulation.




 5   Q15: HOW DO MODELING ASSUMPTIONS RELATE TO REPLICATING POTENTIAL
 6        INEFFICIENCY IN A NETWORK?

 7   A15:   If one were to design a cost modeling process where all technical aspects of a network were

 8          optimized, and where expenses (both capital- and operating-related) reflected forward looking

 9          assumptions, then the resulting cost estimates could approximate those associated with an efficient

10          network. However, as the quotes from the FCC referenced above indicate, some level of network

11          inefficiency may be consistent with the development of economic costs. By allowing wire-center

12          locations to remain “fixed,” based on the incumbent’s existing facilities, the least-cost optimal

13          network design cannot be developed, as would be possible if the so-called “green field” approach

14          were to be utilized.




15          The structure of a cost model will reflect, in part, assumptions the model’s developer makes

16          regarding which network components are fixed and which are allowed to vary and potentially be

17          optimized. Based on these assumptions, the structure of a cost model may be consistent with

18          forward-looking cost principles, or may stray to include embedded costs. Cost models also have

19          user-adjustable inputs. The default user-adjustable inputs will also reflect assumptions made by

20          the model’s designer, however, the user-adjustable inputs may be varied more easily than structural

21          assumptions. Thus, it is important to identify structural assumptions associated with a model, and

                                                        16
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          the assumptions associated with user-adjustable inputs. Both sets of assumptions must be

 2          evaluated in light of the overall objective of developing a cost model which is consistent with

 3          forward-looking cost principles.




 4   Q16: DOES IT APPEAR THAT THERE HAS BEEN A SHIFT IN PERSPECTIVE AT THE FCC
 5        REGARDING THE LEVEL OF FIXED COMPONENTS WHICH ARE ASSOCIATED
 6        WITH A COST MODEL?

 7   A16:   It appears that there may be a shift in perspective, but the ultimate outcome of the pronouncements

 8          associated with the TELRIC NPRM is still unknown. The FCC has indicated that it is considering

 9          whether other network modeling assumptions are better associated with developing cost estimates,

10          especially those which will incorporate more “real world” network attributes.11 However, whether

11          or how the FCC will act on this tentative conclusion is, at this point, highly speculative. Even if the

12          FCC does make significant changes, these will likely be challenged through the appeals process.

13          This can only lead to the conclusion that the FCC’s existing TELRIC rules, as upheld by the U.S.

14          Supreme Court in the Verizon proceeding, must be adhered to in this proceeding. Thus, given the

15          FCC’s and this Commission existing rules, SBC Ohio should not be allowed to move away from

16          the forward-looking economic cost point of view and to move into rate-of- return oriented

17          embedded cost perspective.




18   Q17: WHAT PROBLEMS WOULD A MOVE TO A RATE-OF-RETURN BASIS FOR UNE
19        COST ESTIMATION INTRODUCE?

20   A17:   I am advised by counsel, and I agree, that such a move would be in the first place unlawful. As




            11
                    TELRIC NPRM, ¶52,

                                                         17
                                                                 Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                            On Behalf of OCC
                                                                                 Case No. 02-1280-TP-UNC
                                                                                               Public Version


 1   was noted by the U.S. Supreme Court in the quote above, the Telecommunications Act had an

 2   objective of reducing the ILEC’s monopoly, and rate-of-return regulation is inconsistent with this

 3   objective. In addition, use of an economic cost standard can help with the proper balancing of the

 4   entry paths utilized by CLECs. The Telecommunications Act established a structure which made

 5   three (3) alternative means of entry available to CLECs: pure facilities-based entry, local service

 6   resale, and the availability of UNEs. The Telecommunications Act does not identify any of these

 7   paths as the preferred path, and all three have their merits from the standpoint of enabling different

 8   market strategies for CLECs. The focus on economic cost is most important from the standpoint

 9   of UNEs, as it is with the UNE approach that the Act specifies that rate-of-return principles cannot

10   apply. If UNE rates are developed using a reasonable approach to estimating economic costs,

11   then new market entrants will be led to make their “build or buy” decision on an economically

12   rational basis. If the entrant can self-provide facilities more cheaply than the incumbent (as

13   indicated by the TELRIC-based price for the UNE), then society is made better off because of the

14   more efficient network facilities which are deployed by the entrant. Alternatively, if the entrant finds

15   the UNE rates to be below the cost of building its own network facilities, then society will be better

16   off for the entrant using the incumbent’s facilities.




17   Furthermore, the balance struck by the Telecommunications Act should be considered. SBC and

18   SBC Ohio have become full service telecommunications providers, with the ability to offer local,

19   long distance, Internet, and video services, i.e., to offer full-service bundles. The existence of the

20   three entry paths, including UNEs and UNE-P, provides a check on RBOC retail market behavior.

21   The existence of resale, UNE and UNE-P alternatives allow companies which would otherwise



                                                   18
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


1          be stand-alone providers of, say, wireless or video, to also offer full service bundles.


2   Q18: COULD YOU PLEASE SUMMARIZE YOUR DISCUSSION?

3   A18:   Yes. Given the provisions of the Telecommunications Act it is essential to use economic cost

4          techniques to develop prices for UNEs. Establishing UNE prices without use of rate-of-return

5          regulation is required by law and basing UNE prices on economic costs provides a reasonable

6          means for enabling efficient competitive entry. CLECs can only make entry decisions which will

7          be socially beneficial if UNE prices are based on economic costs.




                                                       19
                                    TAB 2

Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1   Tab 2: Dr. Aron, UNE Rates, Rate-of-Return Regulation and Costs

 2   Q19: WHAT PERSPECTIVE DOES SBC OHIO TAKE IN THIS PROCEEDING WITH
 3        REGARD TO UNE PRICES?

 4   A19:   SBC Ohio’s approach to establishing UNE rates falls clearly into the camp of rate-of-return

 5          regulation. Elsewhere in this testimony I discuss various aspects of the structure of SBC Ohio’s

 6          cost model (the LoopCAT model) and of SBC Ohio’s input assumptions associated with that

 7          model. SBC Ohio’s proposal for UNE pricing associated with LoopCAT is complex, and

 8          understanding how the embedded cost approach associated withrate-of-return regulationmanifests

 9          itself in LoopCAT takes analysis and explanation. However, SBC Ohio’s policy witness, Dr.

10          Aron, clearly identifies the logic behind SBC Ohio’s proposed UNE prices, i.e., application of the

11          principles of rate-of-return regulation and the use of embedded costs. I will turn to a discussion

12          of Dr. Aron’s claims regarding current UNE rates and her criteria for evaluating whether UNE

13          prices are just and reasonable.


14   Q20: DR. ARON INDICATES THAT CURRENT UNE RATES ARE NOT COMPENSATORY. IS
15        HER ANALYSIS CONSISTENT WITH THE PROVISIONS OF THE
16        TELECOMMUNICATIONS ACT?

17   A20:   No. Dr. Aron’s approach is based on the philosophy of rate-of-return regulation. Dr. Aron

18          develops an analysis which compares current UNE revenues with “costs.” However, the costs to

19          which Dr. Aron points are the costs associated with SBC Ohio’s historical books and records.

20          She refers to these as “actual costs,” however, actual costs as used by Dr. Aron begin and end

21          with the Company’s historical books and records, thus they have no bearing on the answer to the

22          question of what the economic costs of the corresponding UNEs might be.




                                                       20
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1   Q21: CAN SBC OHIO’S “ACTUAL COSTS” PROVIDE A FOUNDATION                                                   FOR
 2        EVALUATING THE VALIDITY OF TELRIC-BASED UNE PRICES?

 3   A21:   No. Dr. Aron argues that TELRIC is associated with the incumbent carrier’s “actual costs” and

 4          that the use of actual costs can provide a “reality check” on TELRIC.12 The FCC does not

 5          indicate that the forward-looking costs associated with UNEs and interconnection must be those

 6          that the incumbent is actually incurring. As was discussed earlier, the FCC’s TELRIC rules

 7          develop costs associated with an ILECs network under forward-looking network design and input

 8          price assumptions. As a result, TELRIC can develop the incumbent’s actual forward-looking

 9          costs, which are not related to the incumbent’s actual book costs. Thus, the use of the terminology

10          “actual costs” by Dr. Aron must be recognized in the context of SBC Ohio’s historical books and

11          records, not in terms of economic costs.


12   Q22: AS NOTED BY DR. ARON, DOES THE FCC’S LOCAL COMPETITION ORDER
13        PRECLUDE STATE COMMISSIONS FROM USING ACTUAL COSTS AS A WAY OF
14        TESTING THE VALIDITY OF FORWARD-LOOKING COST ANALYSIS?13

15   A22:   As was discussed above, the FCC’s Local Competition Order recognized the impermissibility

16          of historical costs and developed the TELRIC approach to price UNEs. The inappropriateness

17          of historical costs was reiterated by the FCC in its recent TELRIC rulemaking.

18                  In addition to the problems associated with reliance on incumbent LEC accounting
19                  records, the use of historical costs does not necessarily provide efficient investment signals
20                  to potential entrants. As many economists have noted, it is forward-looking costs, not
21                  historical costs, that are relevant in setting prices in competitive markets. If historical costs
22                  are higher than the forward-looking costs an entrant would face, setting rates on the basis
23                  of historical cost could result in UNE prices that deter entry generally, or cause entrants
24                  to build their own facilities even when it is inefficient to do so. Conversely, if historical
25                  costs are lower than forward-looking costs, UNE rates based on historical costs might


            12
                    Dr. Aron, direct testimony, p. 22, A. 20.
            13
                    Dr. Aron, direct testimony, p. 25, line 448.

                                                          21
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1                  cause entrants to lease facilities when it was more efficient either to build their own or not
 2                  to enter a particular market.14

 3          Thus, the FCC continues to reject the historical cost standard. The quote above also implicitly

 4          identifies a major problem associated with the use of historical cost as a “validity check,” i.e., there

 5          is no predictable relationship between historical costs and economic costs, they may be above or

 6          below economic costs, a point to which Dr. Aron readily admits.15


 7   Q23: DR. ARON INDICATES THAT THE SUPREME COURT RULING IN THE VERIZON
 8        CASE DOES NOT PRECLUDE STATE COMMISSIONS FROM USING ACTUAL COSTS
 9        AS A MEANS TO TEST THE VALIDITY OF FORWARD-LOOKING ECONOMIC
10        COSTS. DOES THIS INDICATE THAT ACTUAL COSTS ARE A USEFUL MEASURE
11        TO BENCHMARK TELRIC?

12   A23:   As I will discuss shortly, I do not agree with Dr. Aron’s claim that the Verizon ruling was favorable

13          to the concept of actual costs, in fact the Supreme Court specifically rejected their use.

14          Overlooking Dr. Aron’s error for a moment, however, the fact that the Supreme Court did not

15          specifically preclude such a benchmarking is not surprising because its objective was to settle the

16          dispute brought before it, not to micromanage the approaches the state commissions might use to

17          comply with the Supreme Court’s conclusion that TELRIC presented a lawful means of complying

18          with the pricing provisions of the Telecommunications Act.


19   Q24: WHAT DID THE SUPREME COURT INDICATE WITH REGARD TO THE USE OF
20        ACTUAL COST TO ESTABLISH TELRIC PRICES?

21   A24:   In his dissent in the Verizon Ruling, Justice Breyer explained his perception of a lack of balance


            14
                    In the Matter of Review of the Commission’s Rules Regarding the Pricing of
                    Unbundled Network Elements and the Resale of Service by Incumbent Local
                    Exchange Carriers, WC Docket No. 03-173, Notice of Proposed Rulemaking,
                    September 15th, 2003, ¶32.
            15
                    Dr. Aron, direct testimony, p. 32, A. 27.

                                                          22
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          associated with the TELRIC rules. Justice Breyer relied on a 1983 Supreme Court decision,

 2          Motor Vehicle Mfrs. Assn. of United States, Inc. v. State Farm Mut. Automobile Ins. Co. 463

 3          U.S. 29, 56 (1983), as justification for his complaint. The majority responded to Justice Breyer’s

 4          dissent as follows:

 5                  . . .Justice Breyer . . .argue(s) that the FCC’s choice of TELRIC bears no “rational
 6                  connection” to the Act’s deregulatory purpose. . . . State Farm involved review of an
 7                  agency’s “changing its course” as to the interpretation of statute. . . this case, by contrast
 8                  involves the FCC’s first interpretation of a new statute, and so State Farm is inapposite
 9                  to the extent that it may be read as prescribing more searching judicial review under the
10                  circumstances of that case. (Indeed, State Farm may be read to suggest the obverse
11                  conclusion, that the FCC would have had some more explaining to do if it had not changed
12                  its course by favoring TELRIC over forward-looking methodologies tethered to actual
13                  costs, given Congress’s clear intent to depart from past ratesetting statutes in passing the
14                  1996 Act.)16

15          As I will discuss in detail in other sections of this testimony, the Supreme Court’s description of the

16          alternative approach to establishing UNE rates, the one which the Supreme Court identified as

17          being appropriately rejected by the FCC is precisely the approach to establishing UNE rates which

18          Dr. Aron advocates and which is inexorably linked to SBC Ohio’s overall cost modeling

19          approach— i.e., applying forward-looking methodologies tethered to actual costs.


20   Q25: DID THE FCC OFFER ANY RELEVANT OBSERVATIONS ON THE ISSUE OF
21        “ACTUAL COSTS” WHEN BRIEFING THE VERIZON CASE?

22   A25:   Yes. The record of that case produced discussion which is highly relevant to Dr. Aron’s claims

23          regarding the use of an “actual cost” validity check:

24                  The incumbents argue that any reasonable forward-looking methodology would have to
25                  be tied to their “actual” forward-looking costs, as opposed to the forward-looking costs
26                  of a “hypothetical” carrier. But they do not explain what they mean by “actual” forward-
27                  looking costs. By definition, forward-looking costs, in contrast to historical costs recorded



            16
                    Verizon Ruling, p. 30, note 20, underline emphasis added.

                                                         23
                                                                Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                           On Behalf of OCC
                                                                                Case No. 02-1280-TP-UNC
                                                                                              Public Version


 1           in the regulatory books of account, do not replicate actual past outlays. They are instead
 2           “the costs that a carrier would incur in the future.” The costs measured by TELRIC are
 3           nonetheless those of the incumbent itself. Those costs are based, moreover, on actual
 4           prices of equipment that is commercially available today—equipment that carriers are
 5           already using to upgrade and expand their networks.

 6           The incumbents appear to be proposing a methodology based on “actual” cost, in today’s
 7           market, of duplicating “actual” existing networks in all physical particulars—or, stated
 8           differently, the “application of up-to-date prices to out-of-date properties.” Economists,
 9           including those upon whom the incumbents rely, uniformly agree that such a measurement
10           is “economically meaningless.” The FCC considered, but rejected, such an approach as
11           “essentially an embedded [i.e., historical] cost methodology,” whichwould produce “prices
12           for interconnection and unbundled elements that reflect inefficient or obsolete network
13           design and technology.”17


14   Thus, the FCC has explicitly rejected the approach advocated by Dr. Aron, and applied by SBC

15   Ohio in this proceeding, i.e., calculating the cost of duplicating their actual network facilities.

16   According to the FCC, the costs of the incumbent are measured with TELRIC, but they are not

17   the actual costs of duplicating the existing network. Instead, starting from the placement of existing

18   wire centers and then applying efficient network design principles, TELRIC estimates the costs that

19   an incumbent would incur under forward-looking conditions. In describing the prices for the

20   appropriately considered equipment associated with a TELRIC study, the FCC does not identify

21   the embedded base of equipment already deployed by a carrier, but rather the new equipment “that

22   carriers are already using to upgrade and expand their networks.” Thus, the Commission should

23   reject Dr. Aron’s perspective because it would result in the “application of up-to-date prices to

24   out-of-date properties” and produce “prices for interconnection and unbundled elements that

25   reflect inefficient or obsolete network design and technology.”



     17
             Reply Brief of the Petitioners United States and the Federal Communications
             Commission. Verizon Communications, Inc., Et Al. v. Federal Communications
             Commission, Et Al. July 2001, pp. 6-7, citations omitted.

                                                  24
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q26: WHAT IS THE RELATIONSHIP BETWEEN PRICES OF PRODUCTS SOLD AND
 2        BOOKED INVESTMENTS IN COMPETITIVE MARKETS?

 3   A26:   In markets which are subject to competitive forces, booked investments and expenses associated

 4          with those booked investments do not influence prices for products sold at the retail level or inputs

 5          which may be sold at the wholesale level. Absent barriers to entry, new market entrants are free

 6          to choose from the overall mix of available technology, including state-of-the-art technology which

 7          delivers the least-cost means of providing the good or service. To the extent that previous

 8          generations of technology exist as potential inputs which could be selected by the entrant, market

 9          forces will adjust the prices of the less efficient inputs in light of efficient new technologies.

10          The less efficient inputs’ book prices will bear no relationship to their market prices. For example,

11          if an incumbent airline operating a fleet of Boeing 727s were to claim predatory “below cost”

12          pricing on the part of an entrant which operated a fleet of more efficient Airbus A320s based on

13          an evaluation of the incumbent’s costs, it could not make its case—the rivals’ costs, based on

14          efficient technology are what matters from an economic and legalperspective. Similar logic applies

15          to forward-looking economic costs associated with UNEs. The claim that embedded costs should

16          be used to determine the TELRIC-compliance of UNE prices is fallacious. Absent a competitive

17          market for the production of UNEs, it is what forward-looking cost models have to say that

18          matters when determining UNE prices. Absent market forces, use of forward-looking cost models

19          will allow regulators to approve input prices which are consistent with the outcome expected in a

20          competitive market.




21          Dr. Aron indicates that both expenses and investments are best checked by use of their associated




                                                        25
                                                               Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                          On Behalf of OCC
                                                                               Case No. 02-1280-TP-UNC
                                                                                             Public Version


 1   historical or embedded levels.18 I will illustrate the flaws in her argument by example, first

 2   discussing expenses and then turning to investment.




 3   An illustration of the lack of economic relevance and inconsistency with TELRIC of Dr. Aron’s

 4   proposed validation approach with respect to expenses is illustrated by a recent Wall Street

 5   Journal article. The article discusses the retirement of the last Boeing 727 jet from U.S. carriers’

 6   scheduled airline service, which occurred in April of 2003 when Delta Airlines decommissioned

 7   its final 727.19 At one time the Boeing 727 was the state-of-the-art passenger aircraft. However,

 8   over time it was surpassed by other aircraft models.




 9   When considering the application of forward-looking cost methodology to the production of airline

10   flights, an approach similar to TELRIC would establish the least-cost forward-looking means of

11   providing flights, given existing airport locations.20     The inappropriateness of Dr. Aron’s

12   methodology which compares forward-looking expenses with those associated with embedded

13   investment can be seen by considering the following:

14           Between 1964 and 1984, Boeing produced 1,832 727s. But what once was economical



     18
             Dr. Aron, direct testimony. See discussion of methodology in footnote (2) in Table 1,
             p. 9.
     19
             Scott McCartney. “Final flight of Boeing’s 727 by U.S. Carriers is a Milestone,” Wall
             Street Journal, February 12, 2003.
     20
             As will be discussed shortly, it is possible that an efficient carrier might choose a mix of
             aircraft that was not exclusively state-of-the-art. However, competitive market forces
             would not allow the prices for obsolete aircraft to reflect their embedded investment.
             Similarly, forward-looking expenses would not be accurately represented by those
             associated with an incumbent carrier’s embedded fleet of aircraft.

                                                 26
                                                               Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                          On Behalf of OCC
                                                                               Case No. 02-1280-TP-UNC
                                                                                             Public Version


 1           proved to be too expensive over time. Boeing began designing jets with two engines,
 2           instead of three, and two pilots, instead of three. The Boeing 757 had more seats than the
 3           727 and transcontinental range. It was far more economical. The Boeing 737 was smaller
 4           and cheaper to fly, and could profitably serve even smaller cities, or provide frequent
 5           service attractive to business travelers. The 737 became Boeing’s best-selling plane,
 6           eclipsing the 727. And Airbus came up with the A320, a plane with a wider fuselage to
 7           offer a bit more passenger comfort, not to mention faster speed and longer range. A 727
 8           consumes about 1,260 gallons of jet fuel an hour, while an A320, carrying the same
 9           number of seats, burns on 788 gallons an hour.

10           Today, the market has evolved beyond just replacing three-engine, three-pilot jets with
11           two-engine, two-pilot planes. The economics of the industry, and the capabilities of
12           airplanes, now dictate that small 50-seat regional jets fly to many of the very same small
13           cities that saw the 727 as their first commercial passenger jet. Using a 150-seat jet to fly
14           to Amarillo, Texas, seems like a silly proposition these days. Three 50-seat regional airline
15           flights offer better revenue potential to carriers than one 150-seat mainline flight. Believe
16           it or not, the three small-jet flights also offer lower costs in terms of crews and fuel.21


17   Applying Dr. Aron’s logic to expenses associated with the technological and market evolution

18   associated with the Boeing 727 scenario provides a clear example of the economically meaningless

19   nature of expenses associated with embedded technologies. If an incumbent carrier with a fleet

20   of 727 aircraft predicted its labor costs, the costs would be associated with the use of three pilots

21   on each flight. That same carrier predicting fuel usage for each 727 flight would identify 1,260

22   gallons per hour. It is clear that this is a very poor predictor of the expenses associated with the

23   forward-looking technology associated with serving a route. For a new entrant at the time the

24   Airbus A320 came onto the market, forward-looking labor expenses for an entrant could be

25   reduced by as much as 33% as compared to the 727, because the Airbus A320 only requires two

26   pilots.22 Given the lower fuel consumption associated with the Airbus A320, fuel expenses could


     21
             “Final Flight of Boeing’s 727 by U.S. Carriers is a Milestone.” Scott McCartney, Wall
             Street Journal, February 12, 2003.
     22
             This discussion abstracts from issues related to labor cost differentials which might arise
                                                                                           (continued...)

                                                 27
                                                                Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                           On Behalf of OCC
                                                                                Case No. 02-1280-TP-UNC
                                                                                              Public Version


 1   be reduced by as much as 38% compared to the incumbent’s embedded 727 baseline. No new

 2   entrant would base its entry decisions on the assumption that its operating expenses would be

 3   constrained by the same levels of expenses associated with the incumbent’s fleet of 727s. This

 4   example illustrates the problem with Dr. Aron’s proposed validation approach with regard to

 5   expenses. The comparison of actual expense levels with those associated with SBC Ohio’s

 6   existing (embedded) base of operations is economically meaningless and cannot be used to validate

 7   the expense levels associated with TELRIC any more than the expense levels associated with

 8   operating a Boeing 727 can be used to validate the expenses associated with operating an Airbus

 9   A320 or fifty-seat regional jet.




10   Turning to the issue of investment, applying TELRIC principles to the airline example would

11   recognize the fact that a new airline would have the option of purchasing or leasing either new state

12   of the art aircraft or less efficient used aircraft. Market forces associated with the sale or lease of

13   aircraft would not allow the book investment values associated with inefficient aircraft to have any

14   influence on the market price of such aircraft that might be considered for lease or purchase by a

15   new entrant. The FCC addressed this issue in its final brief in the Verizon case:

16           The incumbents assert that TELRIC assumes that a carrier would scrap its existing
17           network and rebuild a new, more efficient one every time an advance in technology occurs.
18           TELRIC assumes no such thing.

19           TELRIC instead rests on the rational economic assumption that, as new, more efficient
20           equipment becomes available, the value of older, less efficient equipment will be affected.
21           To use an example suggested by the incumbents, if new more fuel-efficient aircraft were
22           to become available, airlines would not necessarily respond by immediately replacing their



     22
       (...continued)
              from use of non-union pilots.

                                                  28
                                                                  Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                             On Behalf of OCC
                                                                                  Case No. 02-1280-TP-UNC
                                                                                                Public Version


 1           older, less fuel-efficient aircraft, which could still have a significant useful life. The relevant
 2           point is that the market value of the older aircraft would depend not on what the owner
 3           originally paid for it, but on the cost of continuing to operate it relative to the cost of
 4           acquiring and operating the new equipment. Thus, if an airline were to offer to sell or lease
 5           its older aircraft, the price at which it could do so would be constrained by the cost of
 6           acquiring and operating the new aircraft. Airlines considering whether, or how, to enter
 7           the market would base their decision, in significant part, on that cost. To serve its goals
 8           of promoting competition and efficiency in local telecommunications markets while
 9           providing fair compensation to incumbents, the FCC similarly required state commissions
10           to use the long-run forward-looking costs of providing network elements using the most
11           efficient technology currently available.23

12   This statement by the FCC clearly illustrates why Dr. Aron’s proposal to validate TELRIC prices

13   by way of embedded investments is entirely inconsistent with the most basic economic principles.

14   If market forces did exist for the provision of UNEs, the same constraints would apply to an ILEC

15   as apply to an airline planning on leasing aircraft that was not efficient, i.e., “the market value of the

16   older aircraft would depend not on what the owner originally paid for it.”24 Unlike the market for

17   the lease or purchase of used aircraft, which is subject to some competitive forces, the market for

18   UNEs is not. Given the lack of market forces, regulators must replicate the outcome that would

19   be expected if the market were competitive, i.e., “if an airline were to offer to sell or lease its older

20   aircraft, the price at which it could do so would be constrained by the cost of acquiring and

21   operating the new aircraft.” The use of forward-looking economic costs as mandated with the


     23
             Reply Brief of the Petitioners United States and the Federal Communications
             Commission. Verizon Communications, Inc., Et Al. v. Federal Communications
             Commission, Et Al. July 2001, pp. 7-8, italic emphasis in the original, underlined
             emphasis added, citations omitted.
     24
             These market forces are familiar to consumers as well. The owner of a four-year-old
             personal computer may have paid $1,500 for a machine with a 700 megahertz
             processor and 128 Megabytes of RAM when new in 2000. However, even if the
             machine sat unopened in its box and suffered no wear and tear in the intervening four
             years, the market value is no longer $1,500. Given current PC prices and the advance
             of technology in the intervening period, the 2000 machine is obsolete and would likely
             have a market value less than one-tenth of its original price.

                                                    29
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          FCC’s TELRIC is the only way to replicate this market outcome.




 2          Thus, Dr. Aron’s attempted validation of TELRIC by way of embedded investments and of the

 3          expenses associated with those embedded investments is incorrect and economically meaningless.


 4   Q27: DR. ARON DRAWS SUPPORT FOR HER POSITION ON THE USE OF EMBEDDED
 5        COSTS TO “VALIDATE” TELRIC PRICES FROM A BOOK BY DALE LEHMAN AND
 6        DENNIS WEISMAN. DO YOU HAVE ANY COMMENT?

 7   A27:   Yes. The authors of the book referenced by Dr. Aron are both former employees of SBC. While

 8          I am not familiar with Dr. Lehman’s consulting activity, I am aware that Dr. Weisman continues to

 9          consult for SBC and other ILECs.25 Thus, it is not surprising to find the perspectives that are

10          offered in their book are consistent with SBC Ohio’s position.


11   Q28: DR. ARON INDICATES THAT HER EMBEDDED COST ANALYSIS MIGHT NOT
12        ALWAYS RESULT IN HIGHER COSTS. DOES THIS HAVE ANY BEARING ON HER
13        OVERALL ARGUMENT?

14   A28:   Yes. Dr. Aron argues that “actual costs and forward-looking costs are related, and this

15          relationship allows the use of actual costs as a reality check on otherwise unverifiable models, such

16          as TELRIC.”26 However, this view is contradicted by Dr. Aron herself when she indicates that

17          “there is no a priori reason that forward-looking costs necessarily must be lower (or higher) than

18          the costs that are computed from the actual data that the company submits to the FCC in its




            25
                    Dr. Weisman’s vita is available at: http://www.ksu.edu/economics/weisman/
                    Dr. Lehman’s vita is available at: http://polar.alaskapacific.edu/dlehman/
            26
                    Dr. Aron, direct testimony, p. 22, A. 20.

                                                        30
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          ARMIS reports.”27 If there is no definite relationship between booked costs and forward-looking

 2          costs, then use of booked costs as a “reality check” cannot provide much guidance. A far superior

 3          “reality check” is the comparison of estimates produced by a forward-looking cost model which

 4          is consistent with TELRIC principles, such as the FCC Synthesis Model, as I provide later in this

 5          testimony.


 6   Q29: HAS SBC BEEN HIGHLY SUPPORTIVE OF FORWARD-LOOKING COSTS, BASED ON
 7        “UNVERIFIABLE” ECONOMIC COST MODELS IN OTHER CONTEXTS?

 8   A29:   Yes. SBC Ohio, as well as ILECs in general, have long advocated the use of “unverifiable”

 9          forward-looking cost models to enable price reductions for retail services which are subject to

10          some degree of competition. I am not aware of any clamor from ILECs to use their embedded

11          costs as a reality check for price floor calculations because of “unverifiable” forward-looking

12          economic cost models associated with TSLRIC. Given the high degree of comfort with the use

13          of economic cost models for use in determining minimum permissible retail prices, the sudden

14          concern by SBC Ohio (and ILECs in general) about the inability to “validate” TELRIC models

15          should be viewed with extreme caution. In fact, the LoopCAT model is regularly used by SBC

16          to support price floors for services which rely on local loops.28


17   Q30: YOU INDICATE THAT ACTUAL OR HISTORICAL COSTS SHOULD NOT PLAY A
18        ROLE IN PRICE SETTING FOR UNEs. WHAT ABOUT THE IMPACT ON THE
19        PROFITABILITY OF SBC OHIO’S UNE PRODUCT LINE? IS DR. ARON’S CLAIM
20        THAT AN OPPORTUNITY TO EARN A NORMAL RETURN SHOULD BE REFLECTED
21        IN UNE PRICES VALID?



            27
                    Dr. Aron, direct testimony, p. 32, A. 27.
            28
                    “Loop Cost Analysis Tool (LoopCAT) Cost Tool Documentation,” Version 4.01,
                    January 1, 2004, p. 3.

                                                        31
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1   A30:   No. Once again, Dr. Aron advocates rate-of-return regulation where such regulation is not

 2          appropriate. There is no question that the process of setting prices for UNEs may result in the

 3          profitability of the product line being below other product lines offered by ILECs such as SBC

 4          Ohio. However, the guarantee of profits from the sale of UNEs was not the objective of the

 5          Telecommunications Act,29 and for good reason. As noted by the FCC, it is imperative that UNE

 6          prices reflect costs which are consistent with forward-looking principles. These costs can and

 7          should reflect the risk-adjusted compensation of those who supply capital.            This level of

 8          compensation is consistent with competitive market experience. However, any effort to increase

 9          the price levels for UNEs above those associated with a reasonable forward-looking compensation

10          of capital suppliers, and to make shareholders “whole” based on historical cost standards should

11          be strongly opposed. In the first place, such an effort would result in an outcome consistent with

12          rate-of-return regulation, which is prohibited under the Telecommunications Act. Second, making

13          shareholders “whole” for the UNE product line would, by raising UNE prices, send distorted

14          signals to new market entrants, either preventing entry or leading to inefficient duplication of

15          facilities. Furthermore, SBC Ohio should not be allowed to apply principles of ROR regulation

16          selectively, when it sees an advantage to be gained. ROR regulation has traditionally been applied

17          to the entire regulated operations of the regulated company in question. SBC shareholders do not

18          receive their compensation based on the performance of SBC Ohio’s UNE line of business alone,

19          rather compensation is based on SBC Ohio’s overall operations.




            29
                    Section 251(d) pricing constraints indicate that UNE prices “may” include a
                    “reasonable profit.” I have been advised by counsel that such language does not rise to
                    an assurance the UNE prices must provide reasonable profits.

                                                       32
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          It is also important to keep in mind that the opening of the local exchange market, through the

 2          “deal” provided by the Telecommunications Act, gave RBOCs like SBC the ability to access new

 3          markets, an opportunity of which they have vigorously taken advantage. Now that SBC can offer

 4          integrated packages of local and long distance services, Internet services, and video services it has

 5          gained access to new sources of revenues which have the ability to offset any decline in revenues

 6          in the local exchange market. While SBC Ohio may complain about the application of forward-

 7          looking cost methodology to establish UNE prices, similar complaints are notably absent regarding

 8          the upside benefits provided by the Telecommunications Act.




 9   The Desirability of Facilities-Based Competition

10   Q31: DOES THE TELECOMMUNICATIONS ACT SPECIFY THAT FACILITIES-BASED
11        ENTRY IS SUPERIOR TO UNE-BASED COMPETITION?

12   A31:   The law specifies that facilities-based competition is not the only way that an incumbent LEC like

13          SBC Ohio will experience market entry. The Telecommunications Act combines market forces

14          and regulated prices to determine which entry path makes the most sense–i.e., CLECs face a

15          “build or buy” decision. Economic efficiency associated with entry becomes possible because the

16          Telecommunications Act also prevented the application of rate-of-return regulation to UNE price

17          development. Thus, given this provision of the law, and given the necessity of using forward-

18          looking economic costs to determine UNE prices, the critical issue that must be addressed is how

19          to properly establish forward-looking economic costs, which will be used to set economically

20          rational prices for UNEs. As I have discussed above, the starting point to resolve this issue is not

21          the historical costs of the firm or a reversion of principles of ROR regulation.



                                                        33
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          There is no question that by having direct control over their own facilities a CLEC has the ability

 2          to independently innovate and improve the service offering it makes available to its customers. But

 3          it is also important to consider the technical relations of production when evaluating the potential

 4          for facilities-based competition. If facilities-based competition results in increases in the cost of

 5          service which cannot be overcome by the expected dynamic efficiency gains arising from facilities-

 6          based competition, then UNE-based competition may be desirable.


 7   Q32: HAVE YOU DONE ANY STUDIES TO INVESTIGATE THE DYNAMIC EFFICIENCY
 8        GAINS WHICH WOULD BE NECESSARY TO MAKE FACILITIES-BASED COMPETITION
 9        SUPERIOR?

10   A32:   Yes. To explore this issue I conducted multiple runs of the FCC’s Synthesis Model varying the

11          number of customers to study the impact of a variety of market shares on the cost of providing

12          various components of local exchange service. My basic approach was to identify the efficiency

13          losses associated with the provision of UNEs by multiple providers, which would be reflected in

14          increased total cost of providing service. I used data from SBC California’s operations, however,

15          given the nature of the technology in question, I would expect similar results if I were to reproduce

16          the study for SBC Ohio. I designed the study by density zone, which then enabled me to consider

17          the impact of density and market share on costs. Chart 1, below, shows the impact on total loop

18          costs, as reported by the FCC SM, of moving from a pure monopoly of 2-wire local loop

19          providers to a three-firm oligopoly. The results show that total costs increase across density zones,

20          with lower density zones revealing progressively larger total cost increases. I believe that this

21          provides evidence of the social desirability of use of UNEs to enable local exchange competition,

22          and also shows a continuing entry barrier associated with the local loop.



                                                        34
                                                                                                          Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                                                     On Behalf of OCC
                                                                                                                          Case No. 02-1280-TP-UNC
                                                                                                                                        Public Version




                                             Chart 1: Increase in Loop Cost--Monopoly to 3-Firm Oligopoly


70%



60%



50%



40%



30%



20%



10%



0%
      ZONE1,>10,000   ZONE2,5,000-   ZONE3,2,500-   ZONE4,1,000-   ZONE5,500-999 ZONE6,200-499 ZONE7,100-199   ZONE8, 50-99   ZONE9, 25-49   ZONE10,10-24   ZONE11, 0-9
                         9,999          4,999          2,499

                                                                               Density Zone




          Q33: CAN YOU RELATE CHART 1 TO THE POTENTIAL FOR DYNAMIC BENEFITS
 1             ASSOCIATED WITH FACILITIES-BASED LOCAL EXCHANGE COMPETITION?

 2        A33:         Yes. Chart 1 shows the “hurdle” that the dynamic gains from facilities-based competition would

 3                     have to exceed in order to provide an increase in net social benefits. My analysis indicates that the

 4                     costs of moving to the potential to achieve dynamic benefits associated with facilities-based local

 5                     loop competition comes at a price, i.e., the loss of economies of scale which arises when multiple

 6                     loop providers are present in the marketplace. Chart 1 shows that the dynamic benefits of

 7                     facilities-based local exchange competition face an increasing hurdle as customer density

 8                     decreases. For example, to make society better off, the gains required from dynamic efficiency

 9                     resulting from facilities-based competition associated with the highest density zones are a little less


                                                                                       35
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          than 25% of loop cost levels. In other words, the loss of economies of scale leads to the

 2          requirement that dynamic efficiency gains result in other cost reductions by about 25%. As density

 3          decreases, the hurdle increases, making an overall efficiency improvement more difficult to achieve.

 4          Thus, for market segments which tend to be associated with lower density areas, for example,

 5          residential customers, it may be that UNE-loop-based competition provides a superior alternative

 6          to facilities-based competition.




 7   Q34: DOES THE USE OF UNEs OFFSET FACILITIES-BASED COMPETITION?

 8   A34:   In some cases, that is a possibility, but once again, analysis is required. In the residential market

 9          I believe that the existence of UNEs will not be offsetting facilities-based competition. For

10          example, in the residential marketplace the cost structure of the industry generally indicates one

11          where natural monopoly characteristics continue to be associated with the local loop.




12   Q35: DOES THE FACT THAT RESIDENTIAL CUSTOMERS ARE BEING SERVED BY THE
13        UNE-P ARRANGEMENT MAKE THE USE OF UNE-P AN UNDESIRABLE POLICY
14        OUTCOME?

15   A35:   Evaluation of this issue depends on a number of factors. The fact that some consumers are

16          switching to CLECs which rely on the UNE-P indicates that some benefits must be accruing to

17          those customers, i.e., they find the CLEC “deal” superior to SBC Ohio’s in some manner.

18          However, given my evaluation of CLEC marketing to residential customers, it is clear that the high-

19          end of the market is being targeted. A majority of offers appear tailored for customers who use

20          many advanced services or have a significant long distance bill. As a result, a majority of residential

21          customers do not have competitive alternatives at this time. However, as long as those customers


                                                         36
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          who do not have competitive alternatives are protected to some degree through the regulatory

 2          process, and in Ohio that means the application of a form of price regulation, then those customers

 3          who do not have competitive alternatives are not likely to be utilized as a source of subsidy to

 4          offset the high-end residential market losses realized by SBC Ohio. In fact, SBC and other

 5          RBOCs are finding it necessary to forgo some of the margin associated with high-end customers

 6          by developing price-discounted bundles whichare designed to keep high-end residential customers

 7          loyal, i.e., to not switch to an alternative provider.30 Thus, market forces may lead to the “cream”

 8          of the residential market becoming a bit less rich, which would offer some self-correction

 9          associated with the use of UNE-P.


10   Q36: WHAT ABOUT INVESTMENT INCENTIVES THAT MAY BE HARMED AS A RESULT
11        OF TELRIC-BASED UNE PRICES?

12   A36:   Investors supply capital based on perceptions of relative risk. As is noted by Dr. Aron:

13                  Investors have virtually stopped providing capital to CLECs and reveal only the most tepid
14                  enthusiasm for incumbent local exchange companies (ILECs), and then only when ILECs
15                  announce reductions in capital spending.

16          This quote is telling for a number of reasons. First, it would imply, I believe correctly, that the

17          CLEC sector is viewed by the investment community as being relatively more risky than the ILEC

18          sector, a sentiment shared by other industry analysts:

19                  Precursor now advises wholesale avoidance of the competitive telecom segment
20                  and the equipment players, especially the data and optical segment. The relatively
21                  reliable positive cash flow of Verizon, SBC, and Bell South may only be a relatively safe



            30
                    See, for example, “Costly Growth for the Bells,” BusinessWeek Online, May 7, 2004.

            http://www.businessweek.com/investor/content/may2004/pi2004057_0898_pi041.htm


                                                        37
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1                  haven in this very risky sector.31


 2          Thus, from a competitive perspective, the CLECs are at a significant disadvantage in capital

 3          markets as compared to ILECs. If investors have stopped supplying capital to CLECs, one must

 4          ask why, especially given Dr. Aron’s otherwise rosy assessment of the impact of UNE prices on

 5          the operations of CLECs. If CLECs are able to purchase UNEs below cost, investors are not

 6          recognizing it as a sufficient benefit. Rather, given existing UNE prices, CLECs are still considered

 7          highly risky by the investment community. It is difficult to imagine how applying an embedded cost

 8          standard to UNE rates, as suggested by Dr. Aron, would improve investors’ perceptions of CLEC

 9          market risks. Given that investment funds are of highly limited availability to the CLEC sector, it

10          is not clear how a large increase in UNE rates would be successful in spurring CLEC sector

11          investment.


12   Q37: WHAT ABOUT INVESTMENT INCENTIVES FOR SBC OHIO?

13   A37:   Given the nature of telecommunications facilities, and the recent protection which has been granted

14          ILECs subject to unbundling requirements by the FCC, it is unlikely that SBC Ohio will abandon

15          investment as a result of UNE prices which are based on a forward-looking cost foundation. The

16          UNE product is only one of many products which rely on SBC Ohio facilities. The same facilities

17          that provide wholesale services also provide retail services.




18          Furthermore, the FCC’s Triennial Review Order protected ILECs like SBC Ohio from having



            31
                    Scott Cleland, Precursor Group. “Telecom’s Debt Spiral.” February 5, 2002.
                    Emphasis, and style of emphasis in original.

                                                         38
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          to unbundle “next generation” technologies, such as hybrid fiber-copper loops, fiber to the home,

 2          and even reversed the FCC’s previous line-sharing mandate. While portions of the Triennial

 3          Review Order have been overturned by the D.C. Circuit, the appeals court allowed the FCC’s

 4          approach regarding the unbundling next generation technologies to continue.32 Thus, the current

 5          unbundling requirements hold SBC Ohio investment decisions regarding “next generation”

 6          technologies harmless, leaving SBC Ohio free to make its decisions in these areas without concern

 7          for the unbundling provisions of the Telecommunications Act.




 8   Dr. Aron and Fill Factors

 9   Q38: DR. ARON POINTS TO INFORMATION THAT SHE HAS GATHERED FROM OTHER
10        STATES REGARDING APPROVED FILL FACTORS FOR DISTRIBUTION AND
11        COPPER FEEDER PLANT. DOES THAT INFORMATION SUPPORT SBC OHIO’S
12        PROPOSAL TO USE CURRENT ACTUAL FILL LEVELS?

13   A38:   No, if anything, it shows the extreme nature of SBC Ohio’s proposal to use its current actual fill

14          rates. Of the approved fill factors shown in her Chart 7, which addresses distribution fill, only three

15          other states show lower approved fill levels, 34 of the 38 states shown have higher distribution fill,

16          and 24 of the 38 states show approved distribution fill levels which are at least 10 percentage

17          points higher than the distribution fill level proposed by SBC Ohio in its LoopCAT model.33

18          Likewise, in Dr. Aron’s Chart 8, which discusses copper feeder fill factors indicates that all 32

19          states shown have copper feeder fill factors which are higher than SBC Ohio’s proposed level, and



            32
                    U.S. Court of Appeals for the District of Columbia, USTA v. FCC, No. 00-1012,
                    decided March 2, 2004, p. 46.
            33
                    The 24 states exclude counting the current SBC Ohio fill levels approved by the
                    Commission.

                                                         39
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          that 27 of the 32 states have approved copper feeder fill levels which are at least 10 percentage

 2          points higher than SBC Ohio’s proposed actual fill levels.34 In other words, Dr. Aron’s

 3          comparison shows that SBC Ohio’s proposal to utilize actual current fill rates is entirely inconsistent

 4          with the actions taken by most regulatory agencies around the country as they have established

 5          UNE prices using TELRIC methodology.




 6   Q39: DO YOU AGREE WITH DR. ARON’S POSITION THAT INCENTIVE REGULATION
 7        HAS PLAYED AN IMPORTANT ROLE IN ENSURING THAT SBC OHIO’S ACTUAL
 8        CURRENT FILL FACTORS ARE EFFICIENT?

 9   A39:   While I agree that incentive regulation plans have the potential to improve the efficiencyof regulated

10          firms, there is no guarantee that an incentive regulation plan will lead a firm to operate at maximum

11          efficiency. However, the potential for incentive regulation to have had an impact on SBC Ohio’s

12          actual fill levels has been quite limited. The planning guidelines which have resulted in SBC Ohio’s

13          current actual fill levels associated withoutside plant reflect decisions with regard to network design

14          which were made in the period associated with the AT&T monopoly: “SBC Ohio has used the

15          concept of ultimate demand distribution sizing design since adopting AT&T’s SAC (Serving Area

16          Concept) concept in the early 1970s.”35 “The process that SBC Ohio uses today has been in

17          place since AT&T developed the CSA/SAC approach in the 1970s. The SAC is a fundamental

18          building block that, through time, has remained uniform and constant throughout SBC Ohio’s




            34
                    The 27 states exclude counting the current SBC Ohio fill levels approved by the
                    Commission.
            35
                    Response to OCC 1st Set, Interrogatory 57(b).

                                                         40
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          deployment of its network.”36 It appears that the ultimate sizing rule under which decisions made

 2          regarding outside plant deployment approach an “immutable concept” for SBC Ohio, one that does

 3          not change as regulatory constraints or market conditions change. For example, SBC Ohio

 4          indicates that it does not take into account the potential for CLEC self-provision, for DSL

 5          substitution for second lines, or for SBC Ohio’s potential to offer voice over its own broadband

 6          facilities when utilizing the ultimate demand guideline of 2.25 lines per household.37 For the

 7          incentives associated withalternative regulationto function, it must be the case that decision-makers

 8          within the regulated company have the flexibility to respond to the changed incentives. It is quite

 9          clear from statements made by SBC Ohio witnesses, and information provided through discovery,

10          that the design rules which have led to the very low levels of actual fill in SBC Ohio’s outside plant

11          were founded in the period of rate-or-return regulation, and that the management of the company

12          did not have the discretion to respond to incentives arising from alternative regulation. Thus, I do

13          not believe that Dr. Aron’s argument is valid.




14   Q40: HAS SBC OHIO’S ADHERENCE TO THE 30-YEAR-OLD ULTIMATE SIZING
15        GUIDELINES BEEN CONSISTENT?

16   A40:   No. The immutability of the ultimate sizing guidelines has been advanced only when it is to SBC

17          Ohio’s advantage, as is the case when developing the prices for UNEs. While promoting the

18          immutability of the ultimate sizing guidelines for determining the appropriate fill factors for UNEs,

19          SBC Ohio rejects such an approach when determining price floors for services, increasing fill levels



            36
                    Patrick Hamrock, direct testimony, p. 70, lines 6-9.
            37
                    Response to OCC 1st Set, Interrogatory 58(a)-(c).

                                                         41
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          from their actual levels.38 Thus, SBC Ohio has recognized the inappropriate nature of using fill

 2          factors based on ultimate sizing guideline when calculating economic costs associated with retail

 3          price floors.


 4   Q41: DO YOU AGREE WITH DR. ARON THAT COMPETITION WILL “DECONCENTRATE”
 5        THE INDUSTRY, LEADING TO LOWER LEVELS OF FILL IN THE FUTURE, THUS
 6        MAKING EXISTING CURRENT FILL LEVELS “CONSERVATIVE?”

 7   A41:   No. Dr. Aron’s argument must be taken in the context of her testimony regarding the dearth of

 8          facilities-based competition and CLEC reliance on UNEs. Dr. Aron’s point only makes sense if

 9          there is expansive facilities-based competition. If there is not, a fact to which all evidence points

10          at this time, then shifting market share may reduce SBC Ohio’s retail sales, but it does not reduce

11          utilization. Given the use of UNEs and resale, customers continue to be served by the facilities

12          owned by SBC Ohio.




13   Dr. Aron and Depreciation

14   Q42: DR. ARON EQUATES ECONOMIC LIVES WITH DEPRECIATION LIVES USED FOR
15        FINANCIAL REPORTING PURPOSES.39 IS THIS A CORRECT CHARACTERIZATION?

16   A42:   No. As I discuss in greater detail in the section addressing Dr. Vanston’s testimony, lives

17          associated with financial reporting are associated with Generally Accepted Accounting Principles

18          (GAAP). Financial reporting lives are not based on economic depreciation concepts. Without

19          additional information regarding the economic lives of plant, there is no way to determine whether

20          financial reporting lives reflect economic lives. In my discussion of Dr. Vanston’s attempt to


            38
                    Smallwood Deposition transcript, p. 74.
            39
                    Dr. Aron, Direct Testimony, p. 67, lines 1319-1320.

                                                        42
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          demonstrate the extremely short economic life associated with copper distribution plant, I explain

 2          that his assumptions do not reflect either technical or market reality, thus the financial lives adopted

 3          by SBC Ohio are not supported by economic analysis of factors which will influence those lives.




 4   Q43: DO YOU AGREE WITH DR. ARON’S ARGUMENT THAT THE APPLICATION OF
 5        TELRIC MODELS CAUSES DEPRECIATION LIVES OF PLANT TO BE EXTREMELY
 6        SHORT?

 7   A43:   No. Dr. Aron’s argument is based on the highly questionable assumption that at each iteration of

 8          a TELRIC model that all technologies will be superceded by superior versions, and thus at each

 9          iteration of a TELRIC model, all existing plant is obsolete:

10                  The FCC’s TELRIC assumption that the network consists at all times of the state of the
11                  art technology has the effect of decreasing the estimated costs of facilities. However, if
12                  firms were constantly replacing their entire productive capacities as soon as
13                  technology changed, they would never be able to recover their investments unless the
14                  depreciation lives applied for durable goods were extremely short. For example,
15                  suppose a particular asset is placed today, and two years from now a superior or more
16                  efficient technology becomes available. A TELRIC model would require that the asset be
17                  replaced immediately (and a UNE price review would require that an updated TELRIC
18                  model be applied that reflected this new technology ubiquitously deployed in the
19                  network).40
20

21          The fallacy of Dr. Aron’s argument was addressed by the FCC in its supporting brief in the

22          Verizon case, which I cited above.




23          The fact that TELRIC costs may be lower than embedded costs reflects the estimation of the

24          impact that market forces would have on the value of SBC Ohio UNEs, if competition for the

25          provision of UNEs were present. However, absent competition in the provision of UNEs, the


            40
                    Dr. Aron, direct testimony, p. 69, lines 1357-1366, emphasis added.

                                                          43
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


1          appropriate approach is to utilize depreciation rates which reflect a reasonable assessment of the

2          lives of SBC Ohio’s facilities. As I discuss elsewhere in this testimony, I believe that the lives

3          prescribed by the FCC provide a reasonable approach.


4   Q44: PLEASE SUMMARIZE THIS PORTION OF YOUR TESTIMONY.

5   A44:   Dr. Aron’s advocacy of rate-of-return principles should be rejected by the Commission.

6          Application of her suggested approach would result in UNE rates which are inconsistent with the

7          provisions of the Telecommunications Act of 1996 and which are also inconsistent with economic

8          cost methods. It is economically meaningless to validate the results of an economic cost model

9          using historical costs.




                                                      44
                                    TAB 3

Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1   Tab 3: Evaluation of SBC Ohio’s Cost Model

 2   Q45: WHY HAVE YOU FOCUSED YOUR ANALYSIS ON ANALOG 2-WIRE LOOPS?

 3   A45:   The 2-wire basic loop which is used for mass-market access to the public switched network, is

 4          likely to satisfy the Telecommunication Act’s necessary and impair standards for some time. Even

 5          SBC Ohio’s witness Mr. Smallwood recognizes the difficulty of duplicating the local loop facility.41




 6          Given the importance of basic loop technology to the development of local exchange competition,

 7          it is important to set the prices for local loops in a manner which properly reflects economic cost.

 8          The primary impact of the regulated price set for the UNEs will be on CLEC entry decisions.

 9          CLEC business plans will incorporate UNE prices, which will then influence CLEC decisions to

10          build or buy technology. If UNE prices are set to levels which are above economic costs, then

11          CLECs may decide to build facilities when use of the UNE would have been the socially preferred

12          alternative. However, in other instances, as will be the case with local loops, the self-provision

13          option may not exist, given the persistent natural monopoly characteristics associated with local

14          loop technology.




15          On the other hand, if loop prices are set below economic cost, then CLECs will purchase more

16          UNEs than if prices were set at the economic level. In some cases, it might be the case that loop

17          prices which are set too low will discourage the construction of CLEC facilities, but this would

18          assume that construction of such facilities is not impaired by natural monopoly characteristics of the

19          technology.

            41
                    James Smallwood, direct testimony, p. 4, A10.

                                                         45
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1   Q46: WHAT APPROACH IS MOST WIDELY USED TO DETERMINE FORWARD-LOOKING
 2        ECONOMIC COSTS?

 3   A46:   Forward looking economic costs are estimated most frequently using engineering cost models.

 4          These models should estimate the cost of providing network facilities using forward-looking

 5          assumptions and should be capable of optimizing network design.




 6   Q47: FROM A MECHANICAL STANDPOINT, WHAT ARE DESIRABLE CHARACTERISTICS
 7        OF COST MODELS USED TO DEVELOP UNE RATES?

 8   A47:   There are several desirable features of engineering cost models from a mechanical standpoint. (1)

 9          The cost models should be transparent, the theory and practice associated with their operations

10          should be easy to understand. (2) The cost models often require multiple assumptions be changed

11          to generate alternative results or to perform sensitivity analysis. Thus, it is desirable for models to

12          be “integrated” in their construction. An integrated approach would enable a model’s user to

13          modify assumptions and to then have the ramifications of those assumptions automatically update

14          all relevant input values in the model. An integrated approach to creating a cost model thus makes

15          it less likely that data input errors will emerge. For example, if a cost modeling approach uses one

16          spreadsheet to generate annual charge factors, and then those annual charge factors must be

17          manually entered into another spreadsheet to generate recurring cost estimates, there is the potential

18          for data entry errors which can influence the numerical values of cost model outputs. (3) The cost

19          models should be flexible enough to accommodate alternative perspectives on the appropriate

20          approaches to modeling costs. It is important to have the ability to modify assumptions and

21          perform sensitivity analyses associated with estimating costs. If a model is “hard wired” to the

22          approach used by the model’s creator, then evaluating the model will be either unnecessarily


                                                         46
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          difficult or impossible. (4) A cost model should also report results at a level which is reasonably

 2          associated with the technology in question. For example, local loops provide a connection from

 3          a wire center to a customer’s location. The loop plant is thus logically organized around wire

 4          centers, which makes loop cost estimation by wire center a reasonable expectation. Cost models

 5          should allow the auditing of cost model output at an area no larger than the wire center area, and

 6          the ability to analyze costs within a wire center area may provide additional benefits.


 7   Q48: FROM A MECHANICAL STANDPOINT, DOES THE LOOPCAT MODEL PROVIDE A
 8        REASONABLE COST MODEL?

 9   A48:   No, there are a number of problems with the LoopCAT model which make its use cumbersome

10          and which make performing sensitivity analysis difficult or impossible. The LoopCAT model is not

11          integrated. To perform alternative runs of the model, which might result from the modification of

12          various assumptions, for example, cost of capital or depreciation rates, a separate Excel workbook

13          must first be modified to produce annual charge factors. From there, the annual charge factors

14          must be entered into the LoopCAT model itself. This makes performing alternative runs of the

15          model unnecessarily cumbersome.


16   Q49: HAVE YOU HAD THE OPPORTUNITY TO EXAMINE THE LOOPCAT MODEL PRIOR
17        TO ITS FILING BY SBC OHIO?

18   A49:   Yes. I had the opportunity to evaluate the LoopCAT model, as it was filed in California in support

19          of SBC California’s UNE case. I will discuss differences between the LoopCAT which was filed

20          in California and the LoopCAT which was filed in Ohio below. Some of the changes in the

21          LoopCAT model filed in Ohio represent an improvement over that which was filed in California,

22          however, there is one definite new deficiency for the Ohio LoopCAT—the Ohio model does not


                                                        47
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          report results by wire center, rather it reports results by Zone. SBC Ohio indicates that a

 2          LoopCAT model which produces results by wire center for Ohio has yet to be developed.42 This

 3          is a significant disadvantage of the Ohio LoopCAT model, given that wire-center reporting is

 4          possible and has been used by SBC in other jurisdictions. This absence of information in the Ohio

 5          LoopCAT model introduces an unnecessary lack of transparency in evaluating the LoopCAT

 6          model’s results.




 7   Q50: IS THE LOOPCAT MODEL CAPABLE                          OF     PERFORMING             APPROPRIATE
 8        SENSITIVITY ANALYSES?

 9   A50:   No, it is highly limited in this area. For example, SBC Ohio does not know whether the

10          LoopCAT model is capable of determining the least-cost fiber-copper crossover point for

11          individual feeder routes.43 SBC Ohio does not know whether LoopCAT is capable of developing

12          sensitivity runs which result in optimized cable lengths on feeder and distribution routes.44 SBC

13          Ohio is certain, however, that the LoopCAT model cannot be utilized to optimize cable sizes,

14          which would thus enable the calculation of the least-cost means of placing cable in feeder and

15          distribution routes.45




            42
                    Response to OCC’s 2nd Set, Interrogatory 86.
            43
                    Response to AT&T’s 2nd Set of Data Requests, BFP-4.
            44
                    Response to AT&T’s 2nd Set of Data Requests, BFP-6 and BFP-9.
            45
                    Response to AT&T’s 2nd Set of Data Requests, BFP-5.

                                                       48
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1   Q51: GIVEN THE FCC’S REEVALUATION OF TELRIC METHODOLOGY, ISN’T IT
 2        TRUE THAT THE ABILITY OF A COST MODEL TO OPTIMIZE THE DESIGN OF A
 3        NETWORK, GIVEN EXISTING WIRE CENTER LOCATIONS, IS NO LONGER
 4        IMPORTANT?

 5   A51:   No, the ability to optimize the design of a network, given existing wire center locations is still the

 6          guiding principle associated with FCC rules. Even if the FCC were to modify its rules in a manner

 7          which might allow for other “fixed” assumptions, it is unlikely that the use of network optimization

 8          principles will be entirely abandoned. The Telecommunications Act explicitly prohibits use of rate-

 9          of-return regulation when it comes to establishing prices for UNEs and interconnection. I have

10          been advised by counsel, and I agree, that unless this provision of the law is changed, embedded

11          cost methodology, i.e., the methodology associated withrate-of-return regulation, is not permissible

12          for establishing UNE prices. Because economic costs are forward-looking and focus on least-cost

13          technologies, network optimization is still a valid consideration and one that is likely to survive any

14          changes in the FCC’s TELRIC rules. Furthermore, the use of network optimization principles,

15          principles which are consistent with economic cost estimates, are also critical for the development

16          of economically rational UNE prices, i.e., prices whichsend the correct signal to CLECs regarding

17          their decisions to enter, and by which means to enter a market.




18   Cable Sizing

19   Q52: HOW DOES CABLE SIZING ENTER INTO THE COST MODELING PROCESS?

20   A52:   Understanding a loop cost model’s approach to cable sizing is important to interpreting the

21          methods associated witha cost model. When designing a network, it is appropriate to include some

22          amount of spare capacity. Thus, when making decisions regarding cable sizes, network engineers



                                                         49
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          can appropriately include capacity that will not initially be utilized to meet current levels of demand.

 2          The excess capacity planned to meet needs for testing, repair, and growth can be described as the

 3          administrative fill level.




 4          In engineering cost models, it is typical to utilize cable sizing guidelines to identify the capacity of

 5          cables which are needed to provide an efficient network design and a reasonable level of

 6          administrative spare. However, it is also necessary to account for constraints associated with the

 7          availability of cable sizes when calculating costs based on the optimal cable sizes which have been

 8          identified. Thus, engineering cost models typically reflect assumptions regarding cable sizing to

 9          achieve targeted administrative fill levels, and also make adjustments to reflect the reality of the

10          limited number of cable sizes. This process results in the determination of an achieved fill factor

11          for the portion of the network that has been designed.




12   Q53: CAN YOU PROVIDE AN EXAMPLE OF THE INTERRELATED NATURE OF
13        ADMINISTRATIVE FILL AND ACHIEVED FILL?

14   A53:   Yes. Suppose that a study area has 100 households, which average 1.2 lines each, thus current

15          demand is 120 lines or pairs of cable. Suppose further that for administrative purposes, a total of

16          150 pairs is needed. This would imply an administrative fill of 80% (i.e., 120/150). However, the

17          process of building a network is also constrained by the availability of cable sizes. If we assume

18          a 200 pair cable was the smallest size available, the measurable achieved fill in the study area,

19          based on current demand, is 60% (i.e., 120/200). This outcome of a divergence between

20          administrative fill and achieved fill is associated with cost models such as the FCC’s Synthesis



                                                          50
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          Model. Target administrative fills entered at the front-end of the modeling process result in

 2          achieved fills which are lower, reflecting the limited number of sizes available for cable plant.


 3   Q54: WHAT IS THE CONSEQUENCE OF NOT INCLUDING SUFFICIENT EXCESS
 4        CAPACITY IN NETWORK DESIGN?

 5   A54:   If excess capacity is not designed into a network to meet administrative needs and demand growth,

 6          the result is an inefficient network design, because the resulting network will not be able to meet

 7          demand in an efficient fashion. Additional and unnecessary costs will be incurred in the future if a

 8          network is designed without sufficient excess capacity.




 9          However, the required excess capacity must be carefully considered and must have a reasonable

10          upper limit. Too much capacity built into a network results in inefficient investment, i.e., investment

11          that will never be productive. The resulting outcome will also be, as was the case for a network

12          with too little administrative capacity, additional and unnecessary costs incurred in the future,

13          however, the additional costs will be associated with paying for the extra unproductive capacity

14          over time.


15   Q55: HOW DOES THIS DISCUSSION OF CABLE SIZING RELATE TO FORWARD-
16        LOOKING ECONOMIC COSTS?

17   A55:   Forward looking economic costs should reflect a network based on reasonable assumptions

18          regarding the amount of excess capacity required in a network. A reasonable engineering cost

19          model should allow for the possibility of minimizing the costs of building a network in light of

20          reasonable excess capacity constraints. In order for the costs of building a network to be

21          minimized, optimal cable sizing should be performed by a cost model. For example, the FCC’s


                                                         51
                                                                         Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                    On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                       Public Version


 1          Synthesis Model offers the flexibility to identify target fill levels, which then result in achieved fill

 2          levels lower than the targets. For example, in the Virginia Arbitration Order, the FCC adopted

 3          the distribution fill factors associated with the FCC’s Universal Service Inputs Order46, which

 4          ranged from 50% to 75%, depending on the density of a particular zone. However, given these

 5          design fill levels, the overall achieved fill averaged 52%.47




 6   Q56: DOES THE LOOPCAT MODEL ENABLE THE POSSIBILITY OF NETWORK
 7        OPTIMIZATION?

 8   A57:   No, optimization based on cable sizing is not part of the LoopCAT’s capabilities. The LoopCAT

 9          model’s general function is to develop cost estimates based on SBC Ohio’s embedded technology

10          base. Because of its orientation toward updating SBC Ohio’s existing technology base with current

11          technology prices, SBC Ohio has not developed a cost model which is capable of optimizing cable

12          sizes.48 This is a significant problem with the model. Because it cannot optimally size cable, it is

13          hindered from developing cost estimates which are consistent withforward-looking cost principles.




14   Q58: MR. HAMROCK INDICATES THAT DISTRIBUTION SIZING BASED ON “ULTIMATE
15        DEMAND” WILL “OPTIMIZE USAGE OF FACILITIES AND THUS ENSURE
16        EFFICIENT NETWORK OPERATIONS.” DOESN’T THE COMPANY’S USE OF
17        ULTIMATE DEMAND GENERATE AN OPTIMAL NETWORK?


            46
                    In the Matter of Federal-State Joint Board on Universal Service, Forward-
                    Looking Mechanism for High Cost Support for Non-Rural LECs. CC Docket 96-
                    45, 97-160, Tenth Report and Order, November 2, 1999. (Hereinafter, Inputs
                    Order.)
            47
                    Virginia Arbitration Order, ¶250.
            48
                    Response to OCC 2nd Set, Interrogatory 92. Response to BFP-5, AT&T
                    Communications of Ohio’s Second Set of Data Requests.

                                                          52
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1   A58:   No. It is important to understand that there are two major problems related to Mr. Hamrock’s

 2          discussion of ultimate demand. The first issue is the inextricably “backward-looking” nature of

 3          SBC Ohio’s ultimate sizing approach. The second issue is the fact that Mr. Hamrock’s

 4          recommendations regarding ultimate demand are completely ignored by SBC Ohio in the input

 5          values employed in the LoopCAT model. With regard to the first issue, ultimate sizing is a “design

 6          rule” which can be applied when developing a cost estimate. In general, ultimate sizing means that

 7          network design assumptions rule out the possibility of the exhaustion of facilities over a long time

 8          horizon. Mr. Hamrock states:

 9                  Ultimate sizing guidelines are used to try to accommodate all foreseeable demand in the
10                  distribution network. This is accomplished by sizing cable based on a number of “lines per
11                  living unit” that represent the highest expected usage for any given living unit, thereby
12                  creating capacity for sporadic additional line demand in the distribution area, since all living
13                  units will not need this number of pairs.49

14          Ultimate sizing is a highly speculative exercise because the level of “ultimate demand” will only

15          occur many years in the future. The FCC explicitly rejected this approach in its Inputs Order

16          associated with the Universal Service proceeding:

17                  We also affirm our tentative conclusion that the fill factors selected for use in the federal
18                  mechanism should reflect current demand and not reflect the industry practice of building
19                  distribution plant to meet ultimate demand. As we explained in the Inputs Further Notice,
20                  the fact that industry may build distribution plant sufficient to meet demand for ten or
21                  twenty years does not necessarily suggest that these costs should be supported today by
22                  the federal universal service support mechanism.50

23          Furthermore, given the vintage of the assumptions associated with SBC Ohio’s ultimate demand

24          design rule, the design rule can no longer be considered reasonable.



            49
                    Patrick Hamrock, direct testimony, p. 12.
            50
                    Inputs Order, ¶199.

                                                         53
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q59: WHAT IS THE VINTAGE OF SBC OHIO’S ULTIMATE-DEMAND-BASED SIZING
 2        ASSUMPTIONS?

 3   A59:   SBC Ohio utilizes a design rule of “a minimum of two pairs per living unit, plus one maintenance

 4          spare for every four units, plus miscellaneous lines.”51 This network distribution network design

 5          rule is based on practices which have been in effect since the 1970s. Given advances in

 6          technology, the approach proposed by SBC Ohio is not appropriate. Design rules that are 30

 7          years old do not make sense from a forward-looking perspective because of advances in pair-gain

 8          technology. Consider the following example. Suppose that an upper-income household in 1997

 9          wanted a primary line for voice communications and a second line dedicated to dial-up Internet

10          access. As a result, two pairs of distribution cable would be dedicated to this household. Today,

11          satisfying this household’s communication demand generally does not require two pairs. Rather,

12          an upper-income household would likely subscribe to broadband Internet access using digital

13          subscriber line service (DSL), which would reduce the number of lines needed by one. This gain

14          in efficiency is highly evident in SBC Ohio’s network facilities as the number of DSL lines it has in

15          service has increased by over 60% in the last two years.52 As a general proposition, technology

16          advances have pushed pair-gain technology into the distribution plant, thus making design rules

17          implemented in the early 1970s completely out of sync with forward-looking design associated with

18          distribution cables.




            51
                    Patrick Hamrock, direct testimony, p. 12, A. 22. In response to discovery, SBC Ohio
                    could not identify any average number of “miscellaneous lines” above the 2.25 pairs per
                    living unit design rule identified by Mr. Hamrock. As a result, I take Mr. Hamrock’s
                    design rule to be no more than 2.25 pairs per living unit.
            52
                    Response to OCC First Set, Interrogatory 4.

                                                        54
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1   Q60: WHAT IS THE SECOND ISSUE ASSOCIATED WITH SBC OHIO’S CHOICE OF FILL
 2        FACTORS?

 3   A60:   The design rule proposed by Mr. Hamrock would result in a distribution fill of approximately

 4          44%.53 However, the distribution fill utilized in the LoopCAT runs sponsored by Mr. Smallwood54

 5          is based on a fill associated with SBC Ohio’s embedded plant and is considerably below the 44%

 6          factor supported by Mr. Hamrock’s design rule. In urban areas SBC Ohio specifies distribution

 7          fill values of ***(begin proprietary) ****** (end proprietary)*** for urban areas, ***(begin

 8          proprietary) ****** (end proprietary)*** for suburban areas, and ***(begin proprietary) ******

 9          (end proprietary)*** for rural areas.55 The overall weighted average distribution fill level is

10          ***(begin proprietary) ****** (end proprietary)***. The lower utilization rate SBC Ohio

11          employs in its cost studies are based on current utilization associated with SBC Ohio distribution

12          plant.56 This inappropriately inflates investment values in distribution plant and is not based on

13          either a forward-looking view of utilization rates or SBC Ohio’s own design rule for distribution

14          plant.




            53
                     2.25 pairs per living unit translates into a distribution fill of 44.44%. I.e., 1/(2.25) =
                     .4444
            54
                     While the PreProcessFill worksheet in the file “OH 2w Analog LoopCAT 04-07
                     FEB04.xls” has higher distribution fill values specified for the suburban and rural areas,
                     even the highest of these does not rise to the distribution design rule specified by Mr.
                     Hamrock.
            55
                     Fill values from “OHCurrentFillData2004 (Jan04).xls” which was provided with Mr.
                     Smallwood’s workpapers.
            56
                     See the file “OHCurrentFillData2004 (Jan04).xls”, from which the distribution fill values
                     shown in “OH 2w Analog LoopCAT 04-07 FEB04.xls” can be identified.

                                                          55
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1   Q61: HOW DOES THE LOOPCAT MODEL ADDRESS COPPER FEEDER UTILIZATION
 2        RATES?

 3   A61:   SBC Ohio draws its feeder utilization rates from its feeder cable inventory. Actual feeder fill in

 4          SBC Ohio’s network is ***(begin proprietary) ***** (end proprietary)***. This value is

 5          reflected in the default copper feeder fill factors used in the model. By using actual embedded fill

 6          rates for copper feeder, SBC Ohio’s LoopCAT model generates cost estimates that are

 7          unreasonably high. In its comments before the FCC on the input values for the FCC SM, SBC

 8          indicated that with regard to copper feeder, “fills in excess of 82.5% often result in plant

 9          rearrangement costs in order to effectively utilize the remaining spares and balance spare levels in

10          serving area interfaces.”57 This fill level is well above the level used by SBC Ohio in LoopCAT.

11          SBC’s statement to the FCC indicates that fill rates much higher than those currently used by SBC

12          Ohio for copper feeder in LoopCAT could be applied without resulting in higher costs associated

13          with provisioning spare capacity in the network.




14   Q62: HOW DOES SBC OHIO’S DEFAULT ASSUMPTIONS REGARDING FIBER FILL
15        FACTORS AFFECT THE LOOPCAT MODEL’S OUTPUT?

16   A62:   As is the case with the default assumptions affecting distributionand copper feeder, the default fiber

17          feeder fill factors are unreasonably low and thus increase cost estimates. The default values used

18          by SBC Ohio in its LoopCAT model indicate a fill level associated with fiber feeder of




            57
                    “Comments of SBC Communications Inc.” In the Matter of Federal-state Joint
                    Board on Universal Service. Forward-Looking Mechanism For High Cost
                    Support for Non-Rural LECs. CC Docket No. 96-45, CC Docket No. 97-160.
                    July 23, 1999. Page 10.

                                                         56
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          approximately ***(begin proprietary) *** (end proprietary)***.58




 2   Q63: HOW DO THE DEFAULT ASSUMPTIONS IN THE LOOPCAT MODEL ADDRESS
 3        RELATIVE FILL LEVELS ACROSS DENSITY ZONES?

 4   A63:   The fill levels across density zones are handled in LoopCAT in a highly counter-intuitive manner.

 5          It is logical that areas which had higher customer densities would experience higher utilization rates.

 6          Given the fixed cable sizes associated with outside plant, the low density areas would be more

 7          likely to have lower achieved fills. The FCC’s Input Order specified lower fills for copper cable

 8          for lower density areas.59 Furthermore, SBC supported this view in its comments to the FCC on

 9          this issue. SBC agreed with the FCC that “fill factors for copper cable should be lower in the

10          lowest density zones to ensure there will be enough spares available.”60




11          However, this is not reflected in SBC Ohio cost modeling. The LoopCat model utilizes default fill

12          rates for feeder and distribution which are lowest for urban zones. Suburban zones have default

13          values that are lower than rural zones, thus rural zones have the highest fill rates in the LoopCAT

14          model for basic 2-wire loops. As has been discussed by the FCC, the relationship between

15          density and fill factors is appropriately treated when lower density areas have lower fill factors.




            58
                    Fiber fill is from “OH 2w Analog LoopCAT 04-07 FEB04.xls” which was provided
                    with Mr. Smallwood’s workpapers.
            59
                    Inputs Order, ¶193.
            60
                    “Comments of SBC Communications Inc.” In the Matter of Federal-state Joint
                    Board on Universal Service. Forward-Looking Mechanism For High Cost
                    Support for Non-Rural LECs. CC Docket No. 96-45, CC Docket No. 97-160.
                    July 23, 1999. Page 9.

                                                         57
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          Lower fill factors in lower density areas help to account for the use of smaller cables in lower

 2          density areas and for the higher presence of buried cable.61 Thus, the structure of fill factors by

 3          density zone present in the LoopCAT model should be rejected by the Commission.




 4   Q64: PLEASE SUMMARIZE YOUR DISCUSSION OF NETWORK OPTIMIZATION ISSUES
 5        ASSOCIATED WITH THE LOOPCAT MODEL.

 6   A64:   The LoopCAT model does not have the capability to model network costs in a manner consistent

 7          with forward-looking economic costs. Rather, SBC Ohio’s approach starts from the existing

 8          network configuration, which thus pushes the models results toward those associated with

 9          embedded costs. This deficiency ignores Mr. Hamrock’s advice regarding the design of a network

10          to be consistent with “ultimate demand,” which is itself inconsistent with forward-looking cost

11          principles. With regard to copper and fiber feeder, and the distribution portion of plant, SBC

12          Ohio’s application of historical fill factors in the LoopCAT model inappropriately inflates the

13          recurring costs associated with outside plant.




14          Mr. Hamrock indicates that a planning guideline of 2.25 lines per living unit is appropriate for

15          designing optimal outside plant, however, the LoopCAT model does not follow this guideline. The

16          LoopCAT model is not constrained by the engineering guidelines which Mr. Hamrock identifies

17          as optimal. Instead, the LoopCAT model utilizes distribution investment based on historical data

18          associated with the outside plant. As a result, LoopCAT’s modeling outside plant is not forward-

19          looking. The combined impact of SBC Ohio’s assumptions is to increase the adjusted investment



            61
                    Input Order, ¶187.

                                                       58
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          and recurring costs associated with outside plant.




 2   The FCC’s Perspectives on Fill

 3   Q65: HAS THE FCC ADDRESSED THE SPECIFIC LEVEL OF ACHIEVED DISTRIBUTION
 4        FILL WHICH IT VIEWS AS REASONABLE?

 5   A65:   Yes, on several occasions. The FCC’s Input Order62 addressed the inappropriateness of ultimate

 6          demand and its impact on fill factors:

 7                  We find unpersuasive GTE’s assertion that input values for distribution fill factors should
 8                  reflect ultimate demand. In concluding that the fill factors should reflect current demand,
 9                  we recognized that correctly forecasting ultimate demand is a speculative exercise,
10                  especially because of rapid technological advances in telecommunications. For example,
11                  we note that ultimate demand decreases substantially when computer modem users switch
12                  from dedicated lines serving analog modems to digital subscriber lines where one pair of
13                  copper wire provdes the same function as a voice line and a separate dedicated line.63

14          Also, in the §271 proceeding involving SBC’s Kansas and Oklahoma operations the FCC noted

15          that an ALJ decision in an Oklahoma UNE case had relied on utilization rates for network

16          components based on historical levels.64 In response the FCC stated:

17                  . . . The ALJ used a loop fill factor of 30 percent, and rejected the 50 percent figure
18                  proposed by AT&T. Under TELRIC, we determine what the LRIC would be for an
19                  efficient provider. We find that a fill factor that assumes that more than two-thirds of
20                  capacity is idle for an indefinite time is unreasonably low. By way of comparison, the


            62
                    In the Matter of Federal-State Joint Board on Universal Service, Forward-
                    Looking Mechanism for High Cost Support for Non-Rural LECs. CC Docket 96-
                    45, 97-160, Tenth Report and Order, November 2, 1999. (Hereinafter, Input Order.)
            63
                    Inputs Order, ¶200.
            64
                    In the Matter of Joint Application by SBC Communications Inc., Southwestern
                    Bell Telephone Company, and Southwestern Bell Communications Services, Inc.
                    d/b/a Southwestern Bell Long Distance for Provision of In-region, InterLATA
                    Services in Kansas and Oklahoma. CC Docket 00-217. FCC 01-29.
                    Memorandum Opinion and Order, January 21, 2001, ¶79. (Hereinafter, Kansas 271).

                                                       59
                                                                Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                           On Behalf of OCC
                                                                                Case No. 02-1280-TP-UNC
                                                                                              Public Version


 1           Commission adopted fill factors ranging from 50 to 75 percent for the Universal Service
 2           Fund (USF) cost model, the Kansas Commission adopted a 53 percent fill factor for
 3           distribution cable, and the New York Public Service Commission adopted a 50 percent
 4           fill factor. The ALJ’s decision violates TELRIC because it used current fill, and refused
 5           to consider the forward-looking fill or assume that the fill factor would increase over time.65

 6   Thus, the FCC explicitly rejected the Oklahoma ALJ’s reliance on historical utilization rates,

 7   indicating that current fill is not consistent with TELRIC principles.




 8   Most recently in the Virginia Arbitration Order, the FCC Staff specifically rejected Verizon’s

 9   attempt to employ the ultimate demand approach to establishing loop fill factors.

10           In the Inputs Order, the Commission expressly rejected using ultimate demand, as Verizon
11           proposed then and proposes again now, in favor of using current demand to calculate fill
12           factors. There, the Commission found forecasting ultimate demand too speculative. Here,
13           Verizon fails to respond to this concern and provide a method of reliably forecasting
14           ultimate demand, particularly in light of rapidly changing technological developments. Just
15           as the Commission found it inappropriate to include in universal service support the costs
16           of building outside plant designed to meet uncertain ten- or twenty-year demand
17           projections, it is inappropriate for AT&T/WorldCom to bear the cost today of building
18           plant for uncertain ultimate demand.66

19   This decision clearly illustrates the issues raised by the ultimate demand approach advocated (but

20   not followed) by SBC Ohio, namely, whether it is appropriate for those firms which purchase

21   UNEs to pay for plant which has been sized based on uncertain ultimate demand projections. As


     65
             Kansas 271, ¶80, footnotes omitted.
     66
             In the Matter of Petition of WorldCom, Inc. Pursuant to Section 252(e)(5) of the
             Communications Act for Preemption of the Jurisdiction of the Virginia State
             Corporation Commission Regarding Interconnection Disputes with Verizon
             Virginia Inc., and for Expedited Arbitration, In the Matter of Petition of
             WorldCom, Inc. Pursuant to Section 252(e)(5) of the Communications Act for
             Preemption of the Jurisdiction of the Virginia State Corporation Commission
             Regarding Interconnection Disputes With Verizon Virginia Inc., CC Docket Nos.
             00-218 and 00-251. Memorandum and Order, August 29, 2003, DA 03-2738, ¶
             254.

                                                  60
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          was mentioned earlier, given technological change, it is likely that the 30-year-old approach to

 2          ultimate demand utilized by SBC Ohio does not result in a reasonable view of what that demand

 3          might be. Furthermore, in the Virginia Arbitration Order referenced above, the FCC accepted

 4          the FCC’s Synthesis Model administrative distribution fill levels, which resulted in achieved fills of

 5          52.5% in distribution plant.67 In the Virginia Arbitration Order, the FCC adopted administrative

 6          copper feeder utilization rates of between 70% and 82.5%.68




 7   Q66: WHAT HAS THE FCC DONE REGARDING FIBER FILLS?

 8   A66:   The FCC noted when addressing fiber fill factors in the Inputs NPRM:

 9                  Because of differences in technology, fiber fill factors typically are higher than copper
10                  feeder fill factors. Standard fiber optic multiplexers operate on four fiber strands: primary
11                  optical transmit, primary optical receive, redundant optical transmit, and redundant optical
12                  receive. In determining appropriate fiber cable sizes, network engineers take into account
13                  this 100 percent redundancy in determining whether excess capacity is needed that would
14                  warrant application of a fill factor. Both the HAI and BCPM models use the standard
15                  practice of providing 100 percent redundancy for fiber and set the default fiber fill factors
16                  at 100 percent. We tentatively conclude that the input value for fiber fill in the federal
17                  mechanism should be 100 percent.69

18          In its comments on the FCC’s proposed fiber feeder fill factor, SBC stated:

19                  Fiber fill factors of 100% can be obtained because they are not currently subject to daily
20                  service order volatility, as is the case with copper cable, and are more easily administered.
21                  This factor would apply only if fiber redundancy is maintained to allow upgrades and




            67
                    Virginia Arbitration Order, ¶250.
            68
                    Virginia Arbitration Order, ¶260.
            69
                    Further Notice of Proposed Rulemaking, FCC 99-120, May 28, 1999, ¶ 102,
                    citations omitted.

                                                         61
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1                  equipment changeouts without disrupting service to the customer.70

 2          Thus, the fill factor contained in the LoopCAT model for fiber feeder is not forward-looking and

 3          is inappropriate. As the quote above indicates, SBC has recognized this fact in proceedings before

 4          the FCC. In the Virginia Arbitration Order the FCC states as follows on fiber fill factors:

 5                  In the Inputs Order the Commission determined that the ability to upgrade the electronics
 6                  on the ends of the fiber sufficiently accounts for growth, churn and administrative functions.
 7                  The Commission thus adopted a 100 percent fiber feeder fill factor. Further, fiber feeder
 8                  cable is normally installed with 100 percent redundancy. That is, for every fiber strand
 9                  installed a separate strand is installed to account for any breakage that occurs. Thus,
10                  breakage is accounted for in a 100 percent fill factor.71




11   Q67: HAVE YOU UTILIZED THE APPROACH TO FILL FACTORS DESCRIBED BY THE
12        FCC?

13   A67:   Yes. I benchmark the performance of the LoopCAT model by using the FCC’s Synthesis Model,

14          a model which is known to be consistent with forward-looking economic costs. I applied

15          administrative fills at the levels specified by the FCC in its Inputs Order, and used by the FCC

16          Staff in the Virginia Arbitration Order. Because the LoopCAT model does not optimize cable

17          sizes, there I employ fill levels which are consistent with the achieved fills associated with the

18          FCC’s Synthesis Model.




19   Fill Factors and Operating Costs



            70
                    “Comments of SBC Communications Inc.” In the Matter of Federal-state Joint
                    Board on Universal Service. Forward-Looking Mechanism For High Cost
                    Support for Non-Rural LECs. CC Docket No. 96-45, CC Docket No. 97-160.
                    July 23, 1999. Page 10.
            71
                    Virginia Arbitration Order, ¶264.

                                                        62
                                                                         Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                    On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                       Public Version


 1   Q68: MR. HAMROCK STATES THAT HIGHER DISTRIBUTION FILL RATES WILL CAUSE
 2        OPERATING COSTS TO INCREASE, DOES HIS ANALYSIS SUPPORT THIS
 3        PROPOSITION?

 4   A68:   No. Mr. Hamrock indicates that modifying existing fill levels will result in higher operating costs

 5          for distribution plant.72 His Attachment PSH-21 purports to show rising operating costs as fill in

 6          binder groups increases. There are several problems with this analysis. First, the fact that

 7          operating costs rise as binder group fill rises does not imply that higher fills are causing the allegedly

 8          higher costs. Other factors beyond fill affect operating costs. For example, the vintage of plant

 9          may also affect operating costs associated with plant. Older plant may be more prone to failure

10          than newer plant regardless of the fill level associated with the plant. Placement of outside plant

11          may also affect operating costs. For example, plant on aerial structures may experience higher

12          operating costs, regardless of fill levels. Even within similarly placed outside plant, operating cost

13          differences may exist due to variations in climate conditions, again regardless of fill levels.

14          Likewise, operating costs may also reflect differences in the quality of personnel who maintain the

15          plant, again regardless of the fill levels. Mr. Hamrock’s juxtaposition of binder group fill and

16          operating cost does not control for any of these factors, thus his conclusion is entirely unsupported.


17   Q69: ARE THERE OTHER PROBLEMS WITH MR. HAMROCK’S ANALYSIS?

18   A69:   Yes. Mr. Hamrock develops an analysis of operating cost per living unit based on binder group

19          fill rates. By focusing on binder groups as the unit of analysis, Mr. Hamrock distorts the

20          appropriate perspective on the relationship between operating costs and distributionfill. Distribution

21          plant is organized around a wire center, thus a wire center level of analysis is superior to an analysis



            72
                    Patrick Hamrock, direct testimony, p. 38, lines 9-23.

                                                           63
                                                                          Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                     On Behalf of OCC
                                                                                          Case No. 02-1280-TP-UNC
                                                                                                        Public Version


 1          based on binder groups. I believe that analyzing distribution operating costs at the wire center level

 2          is a more meaningful analytical approach because the multiple binder groups that are present at a

 3          given wire center will present an environment where the operating dynamic will involve the use of

 4          binder groups with varying fill levels, not uniform fills as is implied by Mr. Hamrock’s analysis.




 5   Q70: DID    YOU    PERFORM ALTERNATIVE STATISTICAL ANALYSES OF THE
 6        RELATIONSHIP BETWEEN      DISTRIBUTION OPERATING   COSTS  AND
 7        DISTRIBUTION FILL LEVELS?

 8   A70:   Yes. I performed two alternative statistical analyses of SBC Ohio’s distribution operating costs

 9          by CLLI™ code. I performed separate Pearson Product Moment correlation tests between

10          distribution fill by wire center and distribution operating expense. One study evaluated the

11          relationship between fill and operating expense per working pair. The other study evaluated the

12          relationship between fill and operating expense per available pair. Mr. Hamrock utilizes an

13          “operating cost per living unit” approach, however, I believe that using working or available pairs

14          is a better “unit” of evaluation than is living units. Calculation of distribution fill does not involve any

15          count of living units. Operating costs are more likely to be affected by the number of working or

16          available pairs, which is more closely related to consumer demand conditions, such as the need to

17          add another line, rather than the number of living units. I believe that focusing on the number of

18          distribution pairs rather than living units makes more sense.




19   Q71: WHAT ARE THE RESULTS OF YOUR STATISTICAL ANALYSIS?

20   A71:   The Pearson correlation test of distribution fill by wire center and distribution operating costs per

21          working pair reveals a weak negative relationship (-0.06345). This indicates that as fill levels rise,


                                                           64
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          distribution operating expense per working pair falls. However, the result is not statistically

 2          significant at the 5% level. In addition, I performed a Pearson Product Moment correlation test

 3          between the available pairs and fill rates reported by SBC Ohio.73 The result indicates a weak

 4          positive relationship (0.1195) between the fill levels and operating expense per available pair, but

 5          the relationship is not statistically significant at 5%. Thus, the data that I have examined indicates

 6          that there is no evidence that higher fill factors result in increases in operating costs, as is alleged

 7          by Mr. Hamrock.




 8   TSLRIC and TELRIC Fills

 9   Q72: SHOULD FILL FACTORS USED BY SBC IN TSLRIC STUDIES HAVE ANY
10        BEARING ON FILL FACTORS USED IN A TELRIC STUDY?

11   A72:   Yes. Both types of studies are designed to estimate economic costs. TSLRIC studies are

12          designed to develop forward-looking cost estimates for services and TELRIC studies are focused

13          on UNEs. However, in calculating the cost of the service or UNE the same underlying

14          technologies are used, e.g., local loops, switching, transport, etc. There may be some differences

15          in the design of the studies, for example, a TELRIC study will include the costs of selling the

16          unbundled technology to CLECs, while the TSLRIC study will include retail marketing costs.

17          However, the basic structure of the TSLRIC and TELRIC approaches should be similar.




18   Q73: HOW DO SBC AND SBC OHIO ADDRESS THE ISSUE OF FILL FACTORS IN TSLRIC
19        STUDIES?



            73
                    Fill data obtained from “OHCurrentFillData2004 (Jan04).xls”, which was provided
                    with the LoopCAT workpapers.

                                                          65
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   A73:   In the LoopCAT study, which is used to support loop UNE prices, SBC has utilized relatively low

 2          fill factors for the feeder and distribution portions of the loop facility. However, when performing

 3          cost studies for retail services to generate price floors for service which share the local loop, SBC

 4          generally employs fill factors which are much higher than those employed in conjunction with the

 5          LoopCAT model in this proceeding. In fact, many of the fill factors which SBC has utilized in its

 6          TSLRIC studies are higher than those that I propose and utilize in this proceeding. In TSLRIC

 7          studies associated with local exchange service which have been conducted in Arkansas,

 8          Connecticut, Oklahoma, Illinois, Indiana, and Ohio, SBC has employed fill factors with values

 9          considerably higher than the values used by SBC Ohio in this proceeding.




10   Q74: HAS SBC OFFERED AN EXPLANATION FOR THE DIFFERENCE BETWEEN LOOP
11        FILL FACTORS IN TSLRIC AND TELRIC STUDIES?

12   A74:   Yes. However, the explanation offered is not supportable. The argument has been offered by Mr.

13          Smallwood that the higher fills in TSLRIC studies are a result of the spare capacity in a TSLRIC

14          study being a shared cost, which must then be allocated to the services that share the local loop.

15          According to this logic, TELRIC studies must have lower fill factors because the spare capacity

16          is a direct cost for UNEs.74




17          The theory that spare capacity is a shared cost for TSLRIC studies and a direct cost for TELRIC

18          studies is a red herring.75 Mr. Smallwood’s argument requires that economic cost studies be run


            74
                    Smallwood Deposition Transcript, p. 74-76.
            75
                    There might be minor differences between TSLRIC and TELRIC estimates due to
                                                                                       (continued...)

                                                        66
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          under two sets of assumptions. The first set of assumptions would apply to a retail service, where

 2          fill factors would be set at high levels, thus reducing the resulting economic cost estimate. The

 3          second set of assumptions would apply to a wholesale service where the fill factors would be set

 4          at low levels, thus increasing the resulting economic cost estimate. The difference between the two

 5          studies, according to Mr. Smallwood, is the cost of excess capacity.                Mr. Smallwood’s

 6          interpretation is rooted in embedded cost analysis and has no bearing on the question of economic

 7          costs. Within any network, as was discussed above, there is one level of excess capacity which

 8          is optimal. There is no theoretical basis which would then support another calculation of economic

 9          cost based on some other fill level. In fact, the justification for lower fills is clearly stated by Mr.

10          Smallwood and Mr. Hamrock in their testimonies regarding LoopCAT and TELRIC, i.e., to

11          recover the actual historic costs of SBC Ohio’s network facilities. Thus, at its core, Mr.

12          Smallwood’s argument is nothing more than a mixing and matching of economic and embedded

13          cost concepts.




14   Q75: PLEASE SUMMARIZE YOUR DISCUSSION OF FILL LEVELS USED IN TSLRIC AND
15        TELRIC STUDIES.

16   A75:   From an economic perspective, the same assumptions associated with fill levels should be applied

17          in a TSLRIC and TELRIC study. Absent this consistency, the potential for an anti-competitive

18          price squeeze emerges. According to SBC Ohio’s flawed logic, there are two appropriate starting

19          points for forward-looking economic cost studies. Such a split methodology will include embedded


            75
              (...continued)
                     variations in wholesale marketing, or internal vs. external provisioning, but these
                     differences would not approach the magnitude of the difference driven by the
                     divergence of retail and wholesale fill factors used by SBC Ohio.

                                                          67
                                                                         Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                    On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                       Public Version


 1          historical costs in its TELRIC estimates, while using economic costing assumptions with TSLRIC,

 2          where lower cost estimates result in increased pricing flexibility which will be beneficial to SBC

 3          Ohio. Thus, SBC Ohio should not be allowed to strategically select fill factors based on the nature

 4          of the study. Forward-looking economic costs for the same network technology do not take on

 5          one value in a retail environment and another in a UNE environment.




 6   Other Issues with LoopCAT and Distribution Plant

 7   Q76: YOU DISCUSSED PROBLEMS WITH LOOPCAT’S DISTRIBUTION PLANT. ARE
 8        THERE OTHER ISSUES WHICH CAUSE LOOPCAT RESULTS TO BE OVERSTATED?

 9   A76:   Yes. In SBC Ohio’s LoopCAT model, distribution investment is based on the length of the

10          distribution plant. However, the length of distribution plant is not directly measured by SBC Ohio,

11          rather it is estimated. SBC Ohio utilizes a two-step estimation process in the development of the

12          length of distribution cable. First, SBC Ohio identifies the longest distribution cable in a distribution

13          area. Second, to estimate the length of distribution cable in the distribution area, SBC Ohio then

14          calculates an “average” distribution length by multiplying the longest distributioncable by ***(begin

15          proprietary) ***** (end proprietary)***.76 In other words, rather than directly tracking the length

16          of distribution cable, an average distribution length is used as a proxy for the length of all of the

17          distribution cable in a distribution area. It is notable that the value used by Mr. Smallwood in Ohio

18          is higher than the value he used in California by 15 percentage points. There are several major

19          problems with this approach beyond the inflation that has occurred between the California and

20          Ohio filings.



            76
                    SBC Ohio Response to OCC First Set, Request for Admission No. 3.

                                                          68
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q77: WHAT ARE THE PROBLEMS WITH SBC OHIO’S APPROACH TO DETERMINING
 2        DISTRIBUTION LENGTH FOR MODELING PURPOSES?

 3   A77:   The first problem relates to the basis for the number. According to Mr. Smallwood, the number

 4          is based on an analysis of distribution length data for the entire Ameritech region.77 Data provided

 5          by SBC Ohio in Mr. Smallwood’s workpapers shows that Ohio-specific information is available

 6          for distribution length, thus relying on aggregate Ameritech region data is inappropriate.




 7          Second, given that data exists for distribution length for SBC Ohio’s operations, determining a

 8          reasonable measure of average distribution length is an empirical question, i.e., the data provided

 9          by SBC Ohio can be queried and averages evaluated. The data provided with Mr. Smallwood’s

10          workpapers includes distribution length for slightly more than 3 million loops in Ohio (this is

11          apparently all loops which are considered by SBC Ohio to have a distribution component). I

12          queried this database by combined area code and central office code records, which generated

13          249 sample areas. I then compared the true arithmetic average as generated from the data with

14          the “average” based on Mr. Smallwood’s method of multiplying the longest loop by his multiplier

15          value. Mr. Smallwood’s method generated a result which was considerably higher than the actual

16          average.78




            77
                    Smallwood deposition, pp. 107-108.
            78
                    In his deposition Mr. Smallwood indicated that the draw from the ARES database
                    contained zero-length distribution. A check of this database reveals this to be the case,
                    however, the database records exclude the 890,000 direct feeder terminations, thus
                    even if the Ohio-specific information from ARES had been used to develop the
                    distribution lengths, it would have excluded a large number of cases where distribution
                    length is zero.

                                                        69
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          Another problem withSBC Ohio’s approach arises from its modeling assumptions associated with

 2          loops that terminate directly off of feeder.

 3   Q78: DOES THE DIRECT FEEDER TERMINATION ISSUE CAUSE COSTS TO INCREASE
 4        DUE TO THE INCLUSION OF INAPPROPRIATE COSTS OF THE
 5        FEEDER/DISTRIBUTION INTERFACE?

 6   A78:   Yes. SBC Ohio has been aware of this problem for some time. The Indiana Utility Regulatory

 7          Commission noted this problem in its recent order:

 8                  It is also uncontested that SBC’s LoopCAT model fails to reflect the fact that not every
 9                  loop is served via a FDI, which is the terminal linking the feeder and distribution portions
10                  of the loop. In fact, some loops, such as those that terminate at large office buildings, are
11                  served directly via feeder facilities, and do not have a FDI. LoopCAT assumes that very
12                  single loop requires multiple terminations (100% premise termination and 100% FDIs).
13                  The Result is that LoopCAT overstates SBC’s costs by including the costs of block
14                  terminals and FDIs that do not exist.79

15          When asked in discovery whether the cost overstatement problem associated with the LoopCAT

16          model identified by the Indiana Commission had been remedied in the Ohio filing, SBC Ohio

17          indicates that it has not.80 This will likely increase the cost of loops, relatively speaking, more in

18          high density urban areas because of the more frequent termination of loops directly off of feeder.




19   Q79: PLEASE SUMMARIZE YOUR DISCUSSION OF THE LOOPCAT MODEL.

20   A79:   The LoopCAT model violates numerous cost modeling principles, if one is interested in developing

21          an estimate of forward-looking economic costs. Some of these deficiencies are structural, such as


            79
                    In the Matter of the Commission Investigation and Generic Proceeding of Rates
                    and Unbundled Network Elements and Collocation for Indiana Bell Telephone
                    Company, Incorporated, d/b/a SBC Indiana Pursuant to the Telecommunications
                    Act of 1996 and Related Indiana Statutes, Cause No. 42393, January 5, 2004, p.
                    43. (Hereinafter, Indiana UNE Order).
            80
                    Response to OCC’s 3rd Set, Interrogatory No.102, see Attachment TRR-4.

                                                           70
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          the model’s inability to optimize cable size, optimize feeder and distribution length, or determine

 2          the optimal copper/feeder crossover point on a feeder-route basis. The model’s inability to

 3          produce estimates on a wire center basis is another deficiency of the model which, as I will discuss

 4          further below, hinders any evaluation of the model’s output in light of cost information which is

 5          available from other cost models. SBC Ohio’s approach compounds these problems by adopting

 6          assumptions related to its embedded network technology which take the cost analysis further away

 7          from an economic costing approach.




 8   Improvements in LoopCAT

 9   Q80: WHAT IMPROVEMENTS DOES THE CURRENT VERSION OF LOOPCAT CONTAIN?

10   A80:   Previous versions of LoopCAT have assumed that all residential customers were housed in single-

11          family detached housing. The version of LoopCAT filed in Ohio includes a provision to distribute

12          residential customers among the types of structures which Census data indicates are utilized by

13          residential customers in Ohio, whichinclude multiple dwelling units (MDUs). LoopCAT’s previous

14          assumption that no MDUs existed would have increased costs.




15          Previous versions of LoopCAT also made unsupportable assumptions regarding the termination

16          of various services. For example, LoopCAT has previously assumed that large percentages of

17          loop types not typically associated with residential usage, such as payphone, ground start, and

18          DS1, terminated at residential locations in the same frequency as was the case for 2-wire analog

19          loops. This assumption would also likely increase the cost estimate for these loop technologies.

20          While these problems have apparently been addressed in the current LoopCAT version, the


                                                        71
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          problems with LoopCAT are still many. LoopCAT’s point of departure for cost modeling

 2          continues to be based on embedded-costs, thus making its output highly suspect.

 3   Q81: MR. SMALLWOOD       INDICATED    DURING HIS DEPOSITION THAT THE
 4        PERCENTAGES OF MDUs USED IN LOOPCAT WERE DEVELOPED ON A
 5        STATEWIDE LEVEL.81 IS THE STATEWIDE APPROACH APPROPRIATE?

 6   A81:   No. It is more likely that MDUs will be present in urban areas. Thus, SBC Ohio’s introduction

 7          of MDUs into its cost modeling, while an improvement over previous iterations of the model, is

 8          likely to overstate loop costs in higher density areas.




 9   Alternative Cost Modeling

10   Q82: WHAT APPROACH HAVE YOU TAKEN TO MODELING COSTS?

11   A82:   When evaluating a cost model like SBC Ohio’s LoopCAT, perspective can be gained by

12          benchmarking the model with another cost model which is known to be consistent with TELRIC

13          principles. If an alternative cost model that is consistent with forward-looking economic cost

14          principles is used, then by applying comparable assumptions across the two models insight can be

15          gained into the impact of the problems associated with LoopCAT. There are two general

16          categories of problems with LoopCAT, one is its basic structure which does not allow for network

17          optimization and which is oriented toward embedded costs. The other is a set of user adjustable

18          inputs which, while being set by SBC Ohio to inappropriate levels, may nonetheless be modified.

19          With regard to those user-adjustable inputs, it is no secret that three types of assumptions have a

20          very large potential impact on the cost calculations done by any cost model. These are: cost of

21          capital (which includes the costs of debt and equity capital as well as the capital structure),

            81
                    Smallwood Deposition Transcript, p. 59.

                                                         72
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          depreciation lives, and fill factors. If these assumptions are unified and applied across models, one

 2          can get an idea of the structural differences between cost models, and identify potential sources of

 3          bias.


 4   Q83: WHICH ALTERNATIVE COST MODEL IS KNOWN TO BE CONSISTENT WITH
 5        ECONOMIC COST PRINCIPLES?

 6   A83:   I am referring to the FCC’s Synthesis Model or FCC SM. In its efforts to determine the

 7          appropriate level of support for universal service, the FCC convened a proceeding to examine the

 8          state of the art in economic cost modeling. The FCC ultimately settled on a “Synthesis Model,”

 9          which incorporated methodologies associated with what was then known as the Hatfield Model

10          (now known as the HAI model),82 the benchmark cost proxy model (BCPM),83 and the FCC

11          Staff’s own efforts at estimating loop costs. The FCC utilizes the switching and interoffice portions

12          of the HAI model in the FCC SM.




13          I have spent a good deal of time investigating and running the FCC’s Synthesis Model (SM). In

14          addition to becoming familiar with the operations of the model, I have also performed extensive

15          sensitivity analysis of the model to explore the model’s behavior and predictions with regard to the

16          underlying economics of the local exchange. I believe that the FCC SM provides a reasonable

17          estimator of the forward-looking costs of providing UNEs. Clearly, this has been the opinion of the

18          FCC, as they have used the FCC SM to develop forward looking costs associated with universal

19          service funding. The FCC has also relied on results of the FCC SM to assess the reasonableness


            82
                    The HAI model has generally been sponsored by IXCs and CLECs.
            83
                    The BCPM was developed and sponsored by RBOCs.

                                                        73
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          of UNE rates within the context of Section 271 proceedings. Recently the FCC’s Wireline

 2          Competition Bureau utilized the FCC SM to develop UNE prices in the Virginia Arbitration

 3          proceeding.




 4   Q84: HAS THE FCC UTILIZED THE FCC SM TO BENCHMARK COST ESTIMATES AND
 5        UNE RATES?

 6   A84:   Yes, the FCC has utilized the FCC SM for benchmarking purposes in evaluating UNE loop costs

 7          in §271 proceedings. The FCC used the FCC SM with its default input values to assess the

 8          reasonableness of UNE loop prices when considering SBC’s 271 application in Kansas and

 9          Oklahoma:

10                 In taking a weighted average of loop rates in Oklahoma and Texas, we find that
11                 Oklahoma’s rates are roughly one-third higher than those in Texas. This difference is
12                 substantial, but by itself does not indicate that Oklahoma’s rates are not at TELRIC-based
13                 levels. We disagree with commenters who argue otherwise, because as the Department
14                 of Justice notes, “[s]uch differences may arise either from differences in costs between
15                 states, or from different judgements–both of which are reasonable–on rate making issues
16                 that are not susceptible to precise determination.”

17                 As we note above, however, we cannot rely on the judgment of the ALJ with respect to
18                 loop costs. Consequently, we must either conclude that the rate differential is based on
19                 cost, or that Oklahoma’s recurring rates are not TELRIC-based. Our USF cost model
20                 provides a reasonable basis for comparing cost differences between states. . . . Using
21                 weighted average of wire-center loop costs, the USF cost model indicates that loop costs
22                 in SWBT’s Oklahoma Study area are roughly 23 percent higher than loop costs in its
23                 Texas study area. We therefore attribute this portion of the differential, roughly two-thirds
24                 of it, to differences in costs.84

25          By applying the FCC SM to determine loop costs, the FCC concluded that loop costs in

26          Oklahoma were 23 percent higher than loop costs in Texas.



            84
                   FCC’s Memorandum Opinion and Order in CC Docket No. 00-217, FCC Order No.
                   01-29, released January 22, 2001, ¶83, 84 footnotes omitted, emphasis added.

                                                       74
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


 1          In its order approving SBC Pacific’s intraLATA bid in California, the FCC relied on output from

 2          the FCC SM to address the reasonableness of California UNE rates. The FCC conducted a

 3          benchmark comparison between rates in effect in Texas and California based on the FCC SM.

 4          Starting from the premise that SBC’s Texas UNE rates are TELRIC compliant, the FCC then

 5          performed a relative comparison of rates and costs for SBC’s California and Texas operations.

 6          The FCC identified the relative cost differences between the two states, as reported by the FCC

 7          SM, and then assessed the reasonableness of existing UNE rates. The FCC found that the UNE

 8          loop rate in California was 30% lower than the UNE loop rate in Texas. The FCC found that the

 9          Synthesis Model showed that the loop cost in California was 14% lower than the Texas loop costs.

10          Thus, as the FCC found that the percentage difference in UNE rates for 2-wire analog loops

11          between the two states was greater than the percentage difference in costs, as calculated by the

12          FCC SM. The FCC concluded that the UNE loop rates passed its benchmark test, with the

13          California rates thus falling within the TELRIC-compliant range.85




14   Q85: CAN THE FCC SM BE USED TO ASSIST WITH BENCHMARKING A COST STUDY
15        WITHIN A STATE?

16   A85:   Yes. While the FCC has applied its Synthesis Model for evaluating the reasonableness of UNE

17          rates across states, the model also provides a reasonable basis for comparing the results of

18          alternative cost studies within as state. Use of the FCC SM as a means of assessing the

19          reasonableness of the LoopCAT model is appropriate and will provide valuable information to the

20          Commission on the appropriate level of TELRIC for SBC Ohio loops.



            85
                    SBC California 271 Order, ¶63-65.

                                                       75
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q86: HOW DID YOU APPROACH THE BENCHMARKING PROCESS?

 2   A86:   Both the FCC SM and the LoopCAT model contain a number of user-adjustable inputs. The

 3          FCC SM’s default input values are the result of the FCC’s investigation of which inputs should be

 4          used to develop forward-looking economic costs, and was based on an extensive record.86 The

 5          LoopCAT model’s default inputs are based on SBC Ohio’s criteria, as discussed in the

 6          Company’s filing. My objective was to unify the most significant assumptions across the two

 7          models. As I have discussed earlier, and as is discussed by Mr. Smallwood, assumptions

 8          associated with cost of capital, depreciation, and fill factors will have, assumption-for-assumption,

 9          the biggest impact on the results of a cost study. Thus, I developed an approach which I believe

10          results in a reasonable unification of these three very important areas. In addition, I also used the

11          same dollar values for significant inputs associated with outside plant such as copper and fiber

12          cable.




13   Q87: DID YOU MAKE ANY MODIFICATIONS TO THE FCC SM?

14   A87:   Yes, I made one modification to the FCC SM which went beyond what might be considered the

15          standard user-adjustable inputs. The FCC SM estimates the number of DS-0 equivalent lines

16          representing residential, switched business, and special access lines. Special access lines are

17          primarily DS-1 and DS-3 grade, however, the FCC SM calculates high-capacity lines on a DS-0

18          equivalent basis, e.g., the FCC SM counts each DS-1 as 24 DS-0 equivalent channels. The FCC



            86
                     For a summary of the proceedings and the FCC’s conclusions, see: In the Matter of
                     Federal-State Joint Board on Universal Service, Forward-Looking Mechanism
                     for High Cost Support for Non-Rural LECs. CC Docket 96-45, 97-160, Tenth
                     Report and Order, November 2, 1999. (I.e., the Input Order.)

                                                        76
                                                                Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                           On Behalf of OCC
                                                                                Case No. 02-1280-TP-UNC
                                                                                              Public Version


 1    SM constructs facilities, based on line count inputs which include the DS-0 high capacity

 2    equivalents. The impact of this modeling approach is a possible understatement of 2-wire analog

 3    loop costs as the model does its calculations based on a higher than-actual number of loops, thus

 4    forcing unrealistic scale economies in the cost modeling process. To prevent this potential problem,

 5    I used the same approach that was recommended by Verizon and adopted by the FCC Staff in

 6    the Virginia Arbitration Order. Specifically, I zero-out the number of special access lines used

 7    in the cost model, and also set the percentage of switched access lines carried on DS-1 and DS-3

 8    to zero. I believe that this approach is a conservative solution to the problem. In addition, I also

 9    matched the number of DS-0s to the level estimated by LoopCAT, i.e., I trued-up line counts

10    across the two models. This will ensure that the level of potential scale economies will be matched

11    across the two models. Table 1, below, reports the direct cost of SBC Ohio loops in the three

12    zones as reported by the FCC SM before and after the adjustment associated with the special

13    access and high-capacity switched access line counts, and the overall line-count true-up. As can

14    be seen, the impact is as theory would predict—a higher cost estimate is produced by the FCC

15    SM as a result of the adjustment I made.

16   Table 1: Impact of Elimination of Special Access and DS-1 and DS-3 Switched
17   Access Lines on FCC SM Output, Line count matched to LoopCAT.
                                   Zone B       Zone C     Zone D Number of DS-0s
18   Baseline FCC SM Results        $4.80        $7.25      $12.06       5,012,996
19   from FCC Web Site.
20   Adjusted Baseline              $6.40        $8.75      $14.46       3,956,761
21   (eliminating Special Access
22   and Switched Access
23   Carried on DS-1 and DS-3,
24   line count true-up).


25    As is noted by the FCC Staff, this approach to correcting the problem results in the elimination of


                                                  77
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          the sharing of facilities between DS-0 and higher capacity lines and thus costs using this approach

 2          to correcting the voice-grade-equivalent issues will be conservative.87




 3   Unifying Assumptions Across the Models

 4   Q88: WHAT ASSUMPTIONS DID YOU UNIFY ACROSS THE FCC SM AND THE LOOPCAT
 5        MODELS?

 6   A88:   I focused on fill factors, cost of capital, depreciation lives, and the cost of inputs such as copper

 7          and fiber cable, as well as other elements of outside plant, such as FDIs and drop cable. The

 8          process of unifying assumptions across the two models will help identify structural differences in

 9          the models and allow the results of the models to be contrasted.




10   Q89: WHAT CHANGES DID YOU MAKE IN THE “BIG THREE” ASSUMPTIONS ACROSS
11        THE TWO MODELS?

12   A89:   As I mentioned above, depreciation lives, capital costs, and fill factors will, assumption-for-

13          assumption, have the biggest impact on a cost model’s output. For purposes of this analysis, I

14          applied the depreciation lives and salvage values specified by the FCC Staff in the Virginia

15          Arbitration Order.88 With regard to capital costs, I used the costs of debt, equity and capital

16          structure specified by OCC witness Dr. Randall Woolridge.89 With regard to fill factors, there are


            87
                    Verizon Arbitration Order, ¶212.
            88
                    Virginia Arbitration Order, ¶211.
            89
                    Other recent capital costs associated with SBC’s operations include those determined
                    by the Indiana Utility Regulatory Commission, which approved a weighted cost of
                    capital of 9.51%, reflecting a cost of debt of 6.23%, a cost of equity of 11.04%, and a
                    capital structure reflecting 31.76% debt. (Indiana UNE Order, p. 76.) The more
                                                                                                (continued...)

                                                        78
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          fundamental differences in the structure of the FCC SM and the LoopCAT model which make

 2          unification of assumptions a bit more complex. As I have discussed earlier in this testimony, the

 3          LoopCAT model does not optimally size cables, making network optimizationusing the LoopCAT

 4          model impossible. The FCC SM does optimally size cables. As a result, the FCC SM requires

 5          that the user specify design fill factors, which when applied in the model, result in achieved fill

 6          factors which are lower than the design fill levels. In the FCC SM I used the default design fill

 7          values. For purposes of unifying fill assumptions between LoopCAT and the FCC SM, I specified

 8          fill levels in the LoopCAT model which were consistent with the achieved fills in the FCC SM. I

 9          used a value of 50% fill for distribution and 70% fill for feeder in LoopCAT. These achieved fill

10          levels are also consistent with the distribution and copper feeder fill factors which have been

11          approved by the majority of state commissions.




12   Q90: WHAT OTHER INPUT VALUES DID YOU UNIFY ACROSS THE TWO MODELS?

13   A90:   In addition to the unification of the “big three” assumptions discussed above, I also unified the

14          prices of various technology inputs across the LoopCAT and FCC SM. The FCC SM contains

15          a default set of input values for items such as copper and fiber cable, NIDs, drop cable, and for

16          Feeder Distribution Interfaces. The default values are based on the record associated with the



            89
              (...continued)
                     recent Illinois proposed decision arrives at a slightly lower weighted cost of capital,
                     8.94%, which reflects a cost of equity capital at 12.44%, a cost of short-term debt of
                     2.84%, and a cost of long-term debt of 5.80%, with a respective weight for these
                     components of 51% equity, 44.22% long-term debt, and 4.78% short-term debt.
                     Administrative Law Judges’ Proposed Order in case 02-0864, Illinois Bell Telephone
                     Company: Filing to increase Unbundled Loop and Nonrecurring Rates. May 6,
                     2004, p. 83 (Hereinafter, Illinois Proposed Decision).

                                                       79
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          FCC’s Inputs Order and reflect nationwide experience. For example, the FCC SM has default

 2          input cost values for copper and fiber cable which are based on the results of a nationwide NRRI

 3          survey of LECs.90




 4          To unify this aspect of the modeling, I applied the proprietary values for these inputs shown as the

 5          defaults in the LoopCAT model, thus both models were run with proprietary SBC Ohio input price

 6          information. I placed values from LoopCAT in the FCC SM input file for copper cable and fiber

 7          optic cable prices, feeder/distribution interface (FDI) prices, drop cable prices, and network

 8          interface devices (NID) prices. Without exception, the LoopCAT input prices were significantly

 9          lower than the FCC SM’s default input prices. As a result, theory would predict that the FCC SM

10          cost estimates should decrease, as opposed to the cost estimates derived with the FCC’s default

11          input prices levels, and this prediction was realized, as will be discussed further below.


12   Q91: WHEN YOU UNIFY THE ASSUMPTIONS AND INPUT VALUES ACROSS THE
13        FCC SM AND THE LOOPCAT MODELS, WHAT DRIVES ANY REMAINING
14        DIFFERENCES IN THE COST ESTIMATES THAT RESULT?

15   A91:   As I discussed earlier in this testimony, my evaluation of the LoopCAT model leads me to conclude

16          that the model is biased toward an embedded cost rather than a forward-looking economic cost

17          approach. The unification of inputs across the two models will allow an evaluation of this

18          hypothesis. When the user-adjustable assumptions of the model are unified, which I have done for

19          this analysis, then differences in cost output values will tend to reflect the structural differences

20          between the LoopCAT and FCC SM models. The comparison can also reveal the possibility of



            90
                    See discussion in the Inputs Order, ¶105-133.

                                                        80
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          other systematic biases in the LoopCAT model.

 2   Q92: WHAT ARE THE RESULTS OF YOUR ANALYSIS?

 3   A92:   I will present the results which show the impact of the unified assumptions on each model

 4          separately, and then show the results of the two models side-by-side. First, I will discuss the

 5          results of applying the unified set of assumptions on the FCC SM. Recall that the adjusted baseline

 6          results for the FCC SM were reported in Table 1, above. The impact of the unified assumptions

 7          on the FCC SM are shown in Table 2, below, for the three UNE zones.91

 8     Table 2: Impact of the “Unified” assumptions on the FCC SM
 9     Row Model Description                     Zone B        Zone C                           Zone D
10      1 Adjusted FCC SM Baseline (from          $6.40         $8.75                           $14.46
            Table 1, above).
11      2 FCC SM Using Proprietary SBC         ********** **********                         **********
            Ohio cable inputs; unified fills,
            depreciation, and capital costs.


12          As I discussed above, the proprietary input values associated with SBC Ohio’s cost model were

13          uniformly lower than those contained in the FCC SM’s defaults. Furthermore, the weighted cost

14          of capital used for the unifying set of assumptions is lower than the default weighted cost of capital

15          used in the FCC SM. On both of these counts theory would predict a decrease in the FCC SM




            91
                    The same treatment of the DS-0 equivalent issue and line-count true-up, discussed
                    above, is applied for all FCC SM runs. The results in Table 2 for the FCC SM are
                    based on information from the 248 wire centers which are common across the two
                    models. The FCC SM files contained two CLLI™ codes which were not contained in
                    the workpapers supporting LoopCAT. LoopCAT appears to be based on a total of
                    252 CLLI™ code areas. In the workpapers supporting LoopCAT two CLLI™
                    codes which were not contained in the FCC SM files were present. See Attachment
                    TRR-2 for further discussion.

                                                         81
                                                                           Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                      On Behalf of OCC
                                                                                           Case No. 02-1280-TP-UNC
                                                                                                         Public Version


 1              cost estimates, the results shown in Table 2 above bears this out.92

 2   Q93: DID YOU PERFORM A SIMILAR ANALYSIS FOR THE LOOPCAT MODEL?

 3   A93:       Yes. Table 3 shows the baseline LoopCAT results as supported by the testimony of Mr.

 4              Smallwood and compares these with the LoopCAT output based on the unified assumptions. The

 5              assumptions that were modified in LoopCAT were focused on the “big three” of depreciation, cost

 6              of capital, and fill factors. However, I also modified the baseline LoopCAT assumption regarding

 7              the percentage of IDLC and UDLC, applying the assumption of 100% IDLC.93 A full discussion

 8              of modifications made to achieve the results shown in Tables 2, 3, and 4 is contained in Attachment

 9              TRR-2.

10      Table 3: Impact of the “Unified” assumptions on the LoopCAT model (MDF
11      Removed)
12      Row Model Description                    Zone B         Zone C       Zone D
13       1 LoopCAT Baseline (as supported *********** *********** ***********
             by Mr. Smallwood’s testimony).

14          2        LoopCAT using unified fills,          **********        ***********        ***********
                     depreciation, and capital costs.

15              Table 3, above, shows that the unifying assumptions cause significant decreases in the LoopCAT

16              cost estimates.




17   Q94: HAVE YOU PREPARED A TABLE WHICH COMPARES THE UNIFIED LOOPCAT AND


                92
                          The depreciation rates used, drawn from the Virginia Arbitration Order, were very
                          close to the original FCC SM defaults, thus their impact on the cost estimate was slight.
                93
                          To assess the impact of using 100% IDLC, a sensitivity run applied to the LoopCAT
                          output using the “unified” assumptions on capital costs, depreciation, and fill factors.
                          The results revealed that changing the LoopCAT’s default assumption to the 100%
                          IDLC assumption caused loop cost estimates to decrease between 10% and 12%,
                          depending on the zone in question.

                                                             82
                                                                         Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                    On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                       Public Version


 1          FCC SM RESULTS?
 2
 3   A94:   Yes. Table 4, below, shows the results of the two models run with the unifying set of assumptions

 4          that I have discussed.

 5     Table 4: Comparison of FCC SM and LoopCAT with Unified Assumptions
 6     Row Model Description                    Zone B      Zone C        Zone D
 7      1 FCC SM Using Proprietary SBC        ********** *********** ***********
            Ohio cable inputs; unified fills,
            depreciation, and capital costs.
 8      2 LoopCAT using unified fills,        ********** **********    ***********
            depreciation, and capital costs.

 9   Q95: WHAT CONCLUSIONS DO YOU DRAW FROM THE COMPARISON SHOWN IN
10        TABLE 4?

11   A95:   There are two main conclusions that I draw from Table 4. The first is that with the unified set of

12          assumptions, the LoopCAT model produces consistently higher results than does the FCC SM.

13          This indicates the impact of the embedded cost approach associated with the LoopCAT model and

14          reveals problems with the LoopCAT which cannot be readily fixed by altering the basic user-

15          adjustable inputs. The second conclusion which I draw from the results is the fact that the

16          difference in the results between LoopCAT and the FCC SM become more pronounced as the

17          density of the zone increases. I believe that the approach used by LoopCAT to model certain

18          aspects of the network are likely contributing to this. In my discussion of LoopCAT I pointed to

19          three factors which would likely lead to the relatively higher cost estimates in the higher density

20          zones. First, because LoopCAT has introduced multiple dwelling units (MDUs) into LoopCAT

21          based on statewide averages, the economies of serving higher density areas, which are more likely

22          to have MDUs, are assigned to lower density areas. This would cause cost estimates to become

23          relatively higher in high density areas. A second issue which is likely contributing to relatively higher



                                                          83
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          cost estimates in higher density areas is the method of estimating the length of distribution cable

 2          used by SBC Ohio. As was discussed earlier, the basis for estimating the length of distribution

 3          cable is not specific to SBC Ohio, and I believe that it does not present a reasonable means of

 4          estimating an average. Finally, the problem associated with the inclusion of FDI costs for loops

 5          which terminate without distribution would also tend to increase, relatively, the cost estimate in

 6          higher density areas.




 7   Q96: COULD THE COMMISSION UTILIZE YOUR RESULTS OF THE FCC SM
 8        DIRECTLY?

 9   A96:   Yes, these could be used, and I would recommend that the FCC SM results shown in Row 1 of

10          Table 4 be used to identify a lower bound on a reasonable range of direct costs associated with

11          UNE loops. The model has been updated with current input values and the line counts have been

12          trued-up to current levels. If the FCC SM has a weakness, it is related to the customer-location

13          files, which are not up-to-date. However, as weaknesses go, I believe that the lack of current

14          customer-location files pales in comparison with the problems with the LoopCAT model as

15          presented by SBC Ohio.




16   Q97: HOW CAN THE COMMISSION UTILIZE THE BENCHMARKING ANALYSIS WHICH
17        YOU HAVE PERFORMED?

18   A97:   Modifications made to LoopCAT should result in changes in its cost estimates which are consistent

19          with the benchmarking I have performed. That is, (1) the modified LoopCAT cost estimates

20          should reflect relatively lower cost levels than the default output values presented by SBC Ohio,

21          and (2) the modified LoopCAT should also produce relatively lower cost estimates in higher


                                                        84
                                                                Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                           On Behalf of OCC
                                                                                Case No. 02-1280-TP-UNC
                                                                                              Public Version


 1     density areas than is the case with the default model presented by SBC Ohio. If no other

 2     modification to the LoopCAT model were to be made, I would suggest that the LoopCAT cost

 3     output should be trued-up based on information shown in Table 5, below:




 4   Table 5: Relative Cost Difference Across Zones for LoopCAT and FCC SM (Each
 5   model run with the unifying set of input values).
 6   Row        Zone Ratio         LoopCat Zone Cost Ratio    FCC SM Zone Cost Ratio
 7     1            C/B                       73.81%                  68.30%
 8     2            B/D                       64.50%                  40.76%
 9     3            C/D                       87.38%                  59.68%

10     Table 5 shows the ratios of costs generated by each model associated with the three zones. The

11     percentage value in Table 5 reflects the relative relationship between costs generated by each

12     model in the three zones. For example, the ratio of costs associated with the FCC SM associated

13     with Zone B and Zone D shows a value of 40.76%. This says that the FCC SM estimates Zone

14     B costs (the highest density zone) that are 40.76% of the level associated with Zone D (the lowest

15     density zone). By examining Row 2 in Table 5 it can be seen that LoopCAT generates cost

16     estimates for the higher density zones which are a much higher percentage than is estimated by the

17     FCC SM. For example, LoopCAT estimates Zone B costs to be 64.50% of the level associated

18     with Zone D. As I mentioned above, this is likely due to flaws associated with LoopCAT.




19     I used information from the FCC SM to perform a true-up. I adjusted the LoopCAT output so

20     that the same cost ratios across zones associated with the FCC SM are achieved for LoopCAT.

21     Table 6, below, shows the application of the approach.

22



                                                  85
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1     Table 6: True-up/Benchmark Example.
 2      Zone       FCC SM                      Adjustment                     Zone           Adjusted
 3      Pair       Zone Cost                                                               LoopCat Cost
                     Ratio                                                                 (Direct Cost)

 4          B/D     40.76%         LoopCAT Zone D cost * (.4076)                B          ***********
 5          C/D     59.68%         LoopCAT Zone D cost * (.5968)                C          ***********
                                             None                               D          ***********


 6           The Zone D cost value in Table 6 is the cost estimate from the LoopCAT run associated with the

 7           unifying set of assumptions, as reported in Table 4. The benchmarking/true-up then applies the

 8           zone ratios from Table 5 to bring the LoopCAT results into line with the relative performance of

 9           the FCC SM. I believe that the results shown in Table 6 provide an upper bound for establishing

10           a reasonable range of direct cost values for 2-wire UNE loops.




11           This approach will also be useful to the Commission if it accepts modifications to the LoopCAT

12           model which may be suggested by other intervenors. Namely, the outcome of modifying

13           LoopCAT should move the relative results to be more in line with what is observed to be the case

14           with a known forward-looking cost model. This means that additional modifications to LoopCAT

15           should result in the difference between the Zone D and Zone C cost estimates and the Zone D and

16           Zone B cost estimates becoming relatively larger than is the case with the default structure of the

17           LoopCAT model.




18   Q98: PLEASE SUMMARIZE THIS PORTION OF YOUR TESTIMONY.

19   A98:    The LoopCAT model is not consistent with forward-looking cost principles. It has numerous

20           problems, including user-adjustable assumptions associated with depreciation rates and fill factors

                                                         86
                                                             Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                        On Behalf of OCC
                                                                             Case No. 02-1280-TP-UNC
                                                                                           Public Version


1   which are not forward-looking. Beyond the user adjustable inputs, the LoopCAT model has

2   structural problems associated with its inability to optimize cable sizing and network design. The

3   LoopCAT model contains problems known to SBC Ohio regarding its inappropriate inclusion of

4   FDIs for loops that terminate directly off of feeder. The LoopCAT model bases its calculations

5   on estimates of distribution cable length which I do not believe are consistent with SBC Ohio’s

6   network. When benchmarked against a model which is known to be consistent with economic cost

7   principles, the LoopCAT model generates considerably higher cost estimates, and also generates

8   cost estimates which are relatively higher in higher density areas. The Commission should not

9   accept the LoopCAT model, as offered by SBC Ohio, as a basis for UNE rates.




                                               87
                                    TAB 4

Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


 1   Tab 4: Shared and Common Costs

 2   Q99: HAS YOUR DISCUSSION OF COST BENCHMARKING AND COST MODELING
 3        ADDRESSED FORWARD-LOOKING SHARED AND COMMON COSTS?

 4   A99:   No. The analysis presented thus far addresses only direct costs.




 5   Q100: HAVE YOU REVIEWED SBC OHIO’S PROPOSAL FOR ADDRESSING SHARED AND
 6         COMMON COSTS AND DO YOU FIND THE COMPANY’S APPROACH
 7         REASONABLE?

 8   A100: I have reviewed SBC Ohio’s approach, as presented in the testimony of Dr. Kent Currie,

 9          however, I do not find his approach to be reasonable.




10   Q101: CAN YOU DEFINE THE TERMS THAT YOU WILL BE USING IN THIS SECTION?

11   A101: Costs can be classified on a basis of causality. Some costs can be uniquely attributed to the

12          production of a specific service or UNE. These costs are classified as direct costs. Other costs

13          cannot be directly assigned. These may be classified as overhead costs, i.e., costs which arise

14          from a firm being in business and which only become avoidable when the firm goes out of business.

15          Alternatively, shared costs arise where multiple products or services are produced using the same

16          input. From a more technical perspective, shared costs may have specific characteristics, which

17          leads economists to classify the costs as either being joint or common.




18          To illustrate the concept of shared and overhead costs, please refer to Chart 2, below. Chart 2

19          shows a situation where a firm produces two products which share facilities in their production.

20          Chart 2 shows that the production of Services A and B result in direct costs that can be uniquely



                                                       88
                                                                   Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                              On Behalf of OCC
                                                                                   Case No. 02-1280-TP-UNC
                                                                                                 Public Version


1        attributed to either service A or B, and shared costs that arise when either A or B are produced,



                                                    Chart 2
                     Service A                                             Service B



                       Direct Costs                                      Direct Costs




                  Shared Costs Associated with Service A or Service B




                                    Common (Overhead) Costs


2        i.e., the shared costs remain even if the firm decides to stop producing either service A or service

3        B. Chart 2 also shows the common (overhead) costs that arise from being in business.




4        In economic cost terms, the TSLRIC of producing, for example, Service A includes only the direct

5        costs of producing Service A. The TSLRIC does not include either the shared or overhead costs.

6        It should be noted that TSLRIC studies are likely to require the analysis of shared and overhead

7        costs to establish price floors for services.




8   Q102: WITH REGARD TO SHARED COSTS, HOW DOES A TELRIC STUDY DIFFER FROM
9         A TSLRIC STUDY?



                                                         89
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1   A102: The focus of a TELRIC study is unbundled network elements rather than network services. As

 2          a result, a TELRIC study will generate direct costs that are a much larger portion of total costs

 3          than is the case with the TSLRIC of a service. The reason for this is that there are fewer shared

 4          costs with a TELRIC study. For example, in a TSLRIC study associated with local exchange

 5          services, the cost of adding a vertical feature to a network that is already providing all other

 6          services, is quite small. The incremental cost is associated with adding switch software and the

 7          additional costs of billing and supporting the service. However, shared costs are a consideration

 8          with vertical features as the service shares the loop with many other services.




 9          When performing an economic cost study for UNEs, there are relatively fewer shared inputs. For

10          example, the local loop network element is a “product” that new market entrants need to purchase.

11          In a TELRIC study of the local loop, the local loop is a direct cost, not a shared cost. Thus,

12          TELRIC results are much closer to the total costs of producing the network elements, and the unit

13          TELRIC is closer to average costs than is the TSLRIC estimate. The FCC recognized this fact

14          in its Local Competition Order.

15                  [N]etwork elements, as we have defined them, largely correspond to distinct network
16                  facilities. Therefore, the amount of joint and common costs that must be allocated
17                  among separate offerings is likely to be much smaller using a TELRIC methodology
18                  rather than a TSLRIC approach that measures the costs of conventional services.94


19   Q103: DO YOU BELIEVE THAT SBC OHIO’S APPROACH TO DEVELOPING SHARED AND
20         COMMON COSTS IS CONSISTENT WITH FORWARD-LOOKING ECONOMIC COST
21         METHODOLOGY?

22   A103: No.     Dr. Currie’s approach develops factors (multipliers) which are applied to direct cost


            94
                    Local Competition Order, ¶678, emphasis added, footnotes omitted.

                                                        90
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          estimates. The factors thus “gross up” the direct cost to represent a total “forward looking” cost

 2          estimate. Unfortunately, the method applied by Dr. Currie does not generate a forward-looking

 3          factor. A separate factor is developed for common overhead costs and for shared costs. The two

 4          factors are then added to together to yield the overall factor by which direct TELRIC costs will be

 5          increased. The basic idea associated with Dr. Currie’s factors is shown below.

 6          Common cost factor:

 7                  Total Common Costs
 8                   Total Direct Costs

 9          Shared Cost Factor:

10                  Wholesale Shared Costs
11                  Wholesale Direct Costs


12   Q104: WHAT ARE THE PROBLEMS WITH DR. CURRIE’S APPROACH?

13   A104: Dr. Currie’s ratios utilize booked costs, based on ARMIS data in the numerators. This is less than

14          desirable from the standpoint that the FCC’s TELRIC rules require that the shared and common

15          costs reflect a forward-looking perspective.95 However, Dr. Currie compounds this general

16          problem with the creation of his denominators. Rather than using pure historical costs to generate

17          the estimates of the “Total Direct Costs” and “Wholesale Direct Costs” which are used to create

18          the Common and Shared Cost factors, Dr. Currie creates a “forward-looking” value, which is

19          associated with these direct costs. To arrive at the denominator, Dr. Currie starts from booked

20          investments out of ARMIS, but then goes on to reduce the values to bring them into line with

21          current input prices. He reduces the book cost values by applying “current cost to book cost



            95
                    Local Competition Order, ¶672.

                                                       91
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                    Case No. 02-1280-TP-UNC
                                                                                                  Public Version


 1          ratios” which rely on Telephone Plant Indexes (TPIs) which track equipment prices over time.96

 2          Given that equipment prices have fallen over time, the impact on the Direct Cost denominators for

 3          both the shared costs and common cost factors is to reduce the book cost values.


 4   Q105: WHAT PROBLEM THEN ARISES?

 5   A105: Given that the numerators do not face a similar adjustment, the reduction in the values in the

 6          denominators increases the overall ratio, causing the Shared and Common cost multipliers to

 7          increase in value. Thus, not only are the Shared and Common cost multipliers developed by Dr.

 8          Currie rooted in backward-looking historical data, but Dr. Currie also selectively applies a

 9          “forward-looking” adjustment which boosts the numerical value of the overall “backward-looking”

10          multipliers. I found that by eliminating the “book cost to current cost” adjustment from the

11          denominator, the value of the combined Shared and Common Cost Factor decreased by about

12          4.5% percentage points, which reduced the overall magnitude of the cost factor by about 22.5%

13          from the level shown in Dr. Currie’s Confidential Attachment KAC-1.


14   Q106: HAS THIS PROBLEM BEEN APPARENT TO OTHER REGULATORY COMMISSIONS
15         WHICH HAVE REVIEWED DR. CURRIE’S APPROACH?

16   A106: Yes. The Indiana Commission commented:

17                  As we discussed above, because SBC has not developed a “forward-looking” numerator,
18                  we cannot entertain the notion of using SBC Indiana’s “forward-looking” denominator.
19                  By calculating its common cost numerator solely upon embedded, historical data, and its
20                  direct cost denominator using a calculation represented as an estimate of “forward-
21                  looking” direct costs, SBC arrives at a mismatched combination of data. SBC’s analysis
22                  must either rely upon a comparison of forward-looking numerator and a forward-looking
23                  denominator or a booked numerator and denominator, to maintain a proper relationship



            96
                    Dr. Kent Currie, direct testimony, pp. 22-23.

                                                       92
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1                   between common costs (numerator) and direct costs (denominator).97

 2           The proposed decision in Illinois notes:

 3                   According to CLECs, SBC’s use of embedded cost data in the numerators of its shared
 4                   and common cost factors and forward-looking data in its denominators results is an
 5                   inappropriate mismatch of data. CLECs recommend that the Commission rely upon the
 6                   approach of using booked, historic data for both the numerator and the denominator. In
 7                   this way, CLECs assert, though the data used are not forward-looking, the ratio of
 8                   common expenses and direct costs can be considered realistically forward-looking. We
 9                   agree with CLECs’ proposal. 98


10   Q107: WHAT DO YOU RECOMMEND REGARDING THE SHARED AND COMMON COST
11         FACTORS?

12   A107: SBC Ohio’s proposal to mismatch historic costs and “forward-looking” costs should be rejected.

13           The cost factors should be developed ideally by use of a forward looking numerator and

14           denominator. If such a ratio cannot be satisfactorily generated, then a ratio of historic-based costs,

15           appropriately adjusted, may provide an alternative.




16   Q108: DO YOU HAVE ANYTHING ELSE TO ADD ON THE SHARED AND COMMON COST
17         FACTORS?

18   A108: Yes. It is important to keep in mind that these factors will ultimately be applied to the direct cost

19           estimates to generate an overall estimate of forward-looking economic costs. My analysis of the

20           LoopCAT model leads me to conclude that there are significant problems with the estimate of

21           direct costs. This results in the LoopCAT’s cost estimates including costs beyond those that can

22           be reasonably considered to be forward looking. Thus, unless adjustments are made to the direct



             97
                     Indiana UNE Order, p. 138.
             98
                     Illinois Proposed Decision, p. 225.

                                                          93
                                                             Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                        On Behalf of OCC
                                                                             Case No. 02-1280-TP-UNC
                                                                                           Public Version


1   costs produced by LoopCAT, even “perfect” forward-looking shared and common cost factors

2   will result in double recovery of historic costs. The double recovery will arise as the Shared and

3   Common cost factors will gross up any embedded costs present in the LoopCAT estimates.

4   Application of the modified Shared and Common cost factors that I suggest in this testimony should

5   be understood to mean that the resulting multipliers should only be applied to forward-looking

6   direct costs.




                                               94
                                    TAB 5

Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Tab 5: Dr. Vanston and Economic Depreciation

 2   Q109: WHAT IS THE PURPOSE OF THIS SECTION OF YOUR TESTIMONY?

 3   A109: An important aspect of establishing an estimate of forward-looking economic costs is determining

 4          the appropriate rate of depreciation. Depreciation is the mechanism by which the network

 5          investment in an asset is recovered over the life of an asset. In my cost analysis, I employed

 6          depreciation rates in the range specified by the FCC as appropriate. SBC Ohio’s witness Dr.

 7          Lawrence Vanston presents an alternative approach to depreciation which I believe is extreme.

 8          Assessing Dr. Vanston’s testimony requires an evaluation of the prospects for technological and

 9          market substitution for SBC Ohio’s network facilities. Thus, in this section I will address the

10          potential impact of technological and market substitution on the economic values of SBC Ohio’s

11          facilities.


12   Q110: HOW SHOULD DEPRECIATION BE ASSESSED IN THE CONTEXT OF A TELRIC
13         STUDY?

14   A110: Depreciation should reflect economic depreciation. 99 Economic depreciation rates reflect the

15          period-by-period change in the market value of an asset. The market value of an asset reflects the

16          net present value of the cash flows which the asset is expected to generate over the balance of its

17          useful life. Accounting depreciation measures, in contrast, do not reflect the change in market value

18          of an asset.




19   Q111: HOW CAN ECONOMIC DEPRECIATION BE EVALUATED?



            99
                      Local Competition Rules, ¶703. Triennial Review Order, ¶671. TELRIC NPRM,
                      ¶12.

                                                        95
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   A111: The market value of assets will provide insight into the degree to which economic depreciation and

 2           accounting depreciation are similar or divergent. For example, in the mid-1990s Motorola invested

 3           approximately $5 billion in a satellite telephone network called Iridium. However, Motorola had

 4           not done sufficient market research regarding who might be willing to purchase the high-priced

 5           service, and it did not anticipate the rapid global build-out of cheaper ground-based cellular

 6           alternatives. Iridium was forced to shut down and a bankruptcy judge oversaw the sale of

 7           Iridium’s assets for $25 million dollars (i.e., about ½ of one cent on the dollar). The fact that

 8           Iridium’s assets had book values well in excess of market value indicates that economic

 9           depreciation occurred at a very rapid rate. The reason why Iridium’s assets depreciated so quickly

10           is because economic depreciation reflects the expectation of the net present value of cash flows

11           associated with the assets. In Iridium’s case, due to high operating expenses and low subscription

12           rates, the expected future cash flows painted a grim picture, thus the market value of Iridium’s

13           assets decreased dramatically. This illustration of economic depreciation shows the potential for

14           a dramatic divergence between accounting depreciation measures and economic depreciation

15           measures. Iridium operated for less than one year, and the useful life of its assets extended well

16           beyond that period.100 In economic terms, the assets depreciated extremely rapidly. Dr. Vanston

17           paints a similar picture with regard to the future economic value of ILEC assets in general, and

18           SBC Ohio assets in particular. However, his assessment does not hold up when assessed in light

19           of the facts.




             100
                     Iridium’s original assets are still in operation today under different management. See,
                     Forbes.com, “The Return of Iridium,” 11-30-01,
                     http://www.forbes.com/2001/11/30/1130tentech.html

                                                        96
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q112: HOW MIGHT ECONOMIC DEPRECIATION BE EVALUATED FOR A TELEPHONE
 2         COMPANY LIKE SBC OHIO?

 3   A112: Quantifying economic depreciation for telephone company assets may be difficult, as telephone

 4          company assets are not frequently sold in the marketplace. However, there is some evidence that

 5          changing technologies and increasing competition have not caused the economic value of telephone

 6          company assets to decline, as compared to the asset’s book values. For example, in 2002

 7          ALLTEL purchased Verizon’s Kentucky study areas for nearly four (4) times the book value.101

 8          SBC recently announced its intention to sell 650,000 access lines in Michigan and Texas. While

 9          this sale has yet to be consummated, SBC indicates that it believes the assets are worth about $1.5

10          billion.102 If SBC’s beliefs are played out, this would result in a premium of slightly over four (4)

11          times book value. Thus, the evidence with regard to ILEC assets points to a situation opposite of

12          that associated with the Iridium case—ILEC book values are depreciating more rapidly than is

13          reflected in market values. This indicates that market values of ILEC assets are not being

14          destroyed by the threat of technological change and competition. This evidence should be kept in

15          mind when considering Dr. Vanston’s claims which generally indicate the need to dramatically

16          reduce depreciation lives to reflect the expected impact of technological change and competition.



            101
                    For purchase price, see “Alltel Agrees to Buy Phone Lines From Verizon in $1.9
                    Billion Deal,”
                    http://interactive.wsj.com/archive/retrieve.cgi?id=SB100456442797802280.djm

                    Book values calculated from ARMIS 43-01, using Verizon South (Kentucky) and
                    Verizon Contel Kentucky.
            102
                    “SBC Statement Regarding Possible Access Line Sale,” Wall Street Journal, March
                    2, 2004. http://online.wsj.com/article_print/0,,BT_CO_20040302_002310,00.html
                    Book values calculated from ARMIS 43-01, based on all SBC access lines,
                    investments, and reserves in Michigan and Texas.

                                                        97
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1   Q113: ABSENT A QUANTIFIED ASSESSMENT OF THE RELATIONSHIP BETWEEN
 2         MARKET VALUES AND BOOK VALUES OF ASSETS, HOW CAN THE
 3         REASONABLENESS OF DEPRECIATION IN LIGHT OF ECONOMIC PRINCIPLES BE
 4         ASSESSED?

 5   A113: One must evaluate technological and competitive influences and apply judgement. Arguments

 6          indicating the rapid obsolescence of facilities, or the widespread and rapid growth of facilities-

 7          based competition, such as those proffered by Dr. Vanston should be carefully examined. In the

 8          discussion that follows below, I will point to factors which refute Dr. Vanston’s claims. As I will

 9          also discuss, the FCC staff has recently considered this issue and concluded that the FCC’s

10          recommended depreciation lives provide a reasonable standard for economic depreciation at this

11          time.103 I agree with the FCC Staff’s recommended approach to the use of depreciation lives in

12          a TELRIC study.




13   Dr. Vanston’s Approach to Depreciation

14   Q114: WHAT IS THE BASIS FOR DR. VANSTON TESTIMONY REGARDING THE
15         APPROPRIATE DEPRECIATION LIVES FOR SBC OHIO FACILITIES?

16   A114: Dr. Vanston’s conclusions are based on an evaluationof substitution for SBC Ohio facilities. Some

17          of this substitution is based on Dr. Vanston’s assessment of the impact of competition on SBC

18          Ohio’s operations. In addition, Dr. Vanston also considers the impact of technological change on

19          the expected life of various technologies, including switching, copper cable, and fiber cable. I

20          believe, however, that Dr. Vanston’s two-pronged approach to justify short depreciation lives

21          focuses on selective interpretation of information regarding technological and market substitution

22          for SBC Ohio’s existing network technology. As a result, I do not believe the Commission should

            103
                    Virginia Arbitration Order, ¶112.

                                                       98
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          assign much weight to Dr. Vanston’s arguments.




 2   Q115: WHAT IS DR. VANSTON’S                   VIEW        WITH      REGARD TO THE IMPACT OF
 3         COMPETITION?

 4   A115: Dr. Vanston’s approach generally paints a picture of competition as having a dramatic impact on

 5          the life of ILEC assets. He indicates that ILECs are losing millions of lines annually,104 and that this

 6          level of competition is expected to grow dramatically in the future.105 He goes on to point to the

 7          activity of CLECs which utilize SBC Ohio facilities, cable television companies which are planning

 8          to offer voice over Internet protocol (VoIP), and wireless providers as competitive alternatives

 9          which are reducing the economic value of SBC Ohio facilities. He then argues that a shortening

10          of plant lives is appropriate, as compared to either the lives last approved by the Commission, or

11          those recommended by the FCC. I will begin by examining Dr. Vanston’s arguments in light of a

12          more balanced evaluation of competitive alternatives currently facing SBC Ohio customers.




13   Q116: WHAT IS THE PREDOMINANT FORM OF COMPETITION WHICH SBC OHIO
14         FACES IN THE LOCAL EXCHANGE MARKET?

15   A116: SBC Ohio’s primary competitive alternative is provided by CLECs which utilize SBC Ohio

16          facilities. This point is emphasized in Dr. Aron’s testimony and is also supported by my analysis

17          of data supplied by SBC Ohio:

18                  “The number of lines served by CLECs has increased, but these additions have been
19                  driven almost entirely by increases in UNE-P lines while facilities-based line additions (self-



            104
                    Dr. Vanston, direct testimony, p. 29 , line 17.
            105
                    Dr. Vanston, direct testimony, p. 31, line 4.

                                                          99
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1                  provisioning and UNE-L lines) have plummeted. . .”106

 2          Data provided by SBC Ohio indicates that facilities-based entry has affected only a small number

 3          of access lines in SBC Ohio’s service territory. Data provided by SBC Ohio reveals that less than

 4          ½ of one percent of residential lines are served on a facilities basis. Overall, facilities-based

 5          competitors serve about 7% of access lines in SBC Ohio’s service territory. However, there is a

 6          significant split in the distribution of CLEC facilities between business and residential customers.107


 7   Q117: IS THERE EVIDENCE THAT THE WHOLESALE BUSINESS MODEL IS EMERGING
 8         AS A COMPONENT OF SUCCESSFUL OPERATIONS IN THE TELE-
 9         COMMUNICATIONS INDUSTRY?

10   A117: While the ILECs are required by law to make their network facilities available on a wholesale

11          basis, other telecommunications providers are adopting this strategy voluntarily. Some companies,

12          such as Level 3 Communications are focused on wholesale markets, providing bandwidth and

13          network services to other firms. However, even retail-oriented telephone companies are beginning

14          to recognize the advantages of establishing themselves in wholesale markets. Sprint, for example,

15          has taken steps to resell its wireline and wireless network services to rivals. Sprint points to the

16          ability to achieve superior scale economies through more intensive use of its facilities as one

17          motivation for using the resale strategy. 108 Thus, the strategic use of wholesale markets is an


            106
                    Dr. Aron, direct testimony, p. 47.
            107
                    Overall market share data as of December 2003, based on “The Status of Competition
                    in Ohio,” provided with SBC Ohio workpapers. Residential customer market share
                    based on the same document, with additional information provided in response to OCC
                    First Set, Interrogatories 10 and 66.
            108
                    “Sprint’s Role as Wholesaler: ‘Arms Dealer’ to the Industry,” Wall Street Journal,
                    May 21, 2004.
                                                                                             (continued...)

                                                         100
                                                                         Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                    On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                       Public Version


 1           expanding fact in the telecommunications industry. A company like Sprint is undertaking the

 2           wholesale strategy because its management believes that its addition of a wholesale approach will

 3           improve the value of its firm.




 4   Q118: WHY DO YOU BELIEVE THAT WHOLESALE BUSINESS MODELS WILL BECOME
 5         INCREASINGLY IMPORTANT IN THE TELECOMMUNICATIONS INDUSTRY?

 6   A118: The focus on bundling of service offerings will provide motivation for telecommunications providers

 7           to become either suppliers or users of wholesale services. Growth in bundle-based competition

 8           will lead those firms which are not well positioned to offer multiple services to seek out the “piece

 9           parts” which are missing from their service offerings. AT&T’s recent announcement regarding its

10           intent to utilize Sprint wireless facilities illustrates an example of this type of strategy. In another

11           case, by selling wireless and wireline services at wholesale to its ostensible rival Time Warner,

12           Sprint gains a place for its technology in Time Warner’s bundle of services, with Time Warner

13           facilities providing video and high-speed Internet access, and Sprint providing the conventional

14           wireline and wireless telecommunications services. Thus, I believe that strategically offering

15           wholesale offerings will be viewed as an important strategy in the coming years, one which will

16           enhance economic values of assets.




17   Q119: DOES UNE-BASED COMPETITION HAVE THE SAME IMPACT ON THE
18         ECONOMIC VALUE OF SBC OHIO’S NETWORK TECHNOLOGY AS FACILITIES-
19         BASED COMPETITION?

20   A119: No. If CLECs utilize ILEC facilities, the economic value of the facility is not affected to the same


             108
                (...continued)
                      http://online.wsj.com/article_print/0,,SB108508486331117175,00.html

                                                          101
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1           degree as is the case when CLECs build their own facilities. Thus, to support his contention that

 2           lives of SBC Ohio facilities must be reduced, Dr. Vanston must emphasize facilities-based

 3           competition, which other SBC Ohio witnesses, as well as data supplied by SBC Ohio, indicate has

 4           not advanced to a very high level in Ohio. Dr. Vanston thus indicates that there will be a significant

 5           impact on SBC Ohio facilities from services other than those provided by CLECs, namely that

 6           consumers have either begun or will shortly begin to migrate from SBC Ohio’s facilities to wireless

 7           and cable TV phone services which will undermine SBC Ohio’s position in the marketplace. I do

 8           not believe that these alternatives are likely to have a significant impact on the economic value of

 9           SBC Ohio’s assets, as is suggested by Dr. Vanston.




10   Wireless Substitution

11   Q120: WHAT IS THE PURPOSE OF THIS SECTION OF YOUR TESTIMONY?

12   A121: Dr. Vanston indicates that wireless telephony will substitute for wireline telephony and thus cause

13           SBC Ohio assets to depreciate rapidly. In this section I will assess the extent to which wireless

14           substitution is possible. I conclude that the evidence points to wireless as providing a benefit to

15           providers such as SBC Ohio, and which will not require shortened depreciation lives.




16   Q122: DO WIRELESS TELEPHONE SERVICES OFFER A COMPETITIVE ALTERNATIVE
17         TO WIRELINE SERVICES THAT REQUIRE THE DRAMATIC REDUCTION IN
18         DEPRECIATION LIVES RECOMMENDED BY DR. VANSTON?

19   A122: No. There is not sufficient evidence to conclude that wireless telephones are a competitive

20           alternative to a land-line telephone which would dramatically affect the economic value of SBC

21           Ohio’s plant. In fact, as I will discuss further below, evidence points to wireless telephony


                                                         102
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          enhancing the economic value of wireline networks, and for integrated providers of wireless and

 2          wireline services, such as SBC, the enhancement in economic value of wireline facilities is likely

 3          significant.




 4   Q123: DO YOU AGREE WITH DR. VANSTON’S OPINION THAT WIRELESS SUBSTITUTION
 5         IS UNDERWAY AND LIKELY TO DRAMATICALLY INCREASE IN THE VERY NEAR
 6         FUTURE?

 7   A123: No. As I will discuss below, there are many obstacles to replacing wireline with wireless

 8          telephony. These obstacles, including price points which make wireless telephony a much more

 9          expensive alternative for most users, will constrain the maximum number of subscribers which will

10          ultimately make the switch to a pure wireless alternative. While it is possible that certain

11          demographic groups, for example, college students, may rely on a wireless phone more frequently,

12          this demographic group was previously less likely to have a telephone in the first place, thus the

13          widespread use of wireless by this demographic certainly does not represent a displacement of

14          wireline telephones.109 More general data on wireless and wireline subscription indicates that

15          wireless telephones are viewed as an economic complement, rather than as an economic substitute

16          for wireline telephones.




17   Q124: ISN’T IT TRUE THAT THE DECLINE OBSERVED IN SBC OHIO ACCESS LINES IN
18         RECENT YEARS IS PARTIALLY ATTRIBUTABLE TO THE GROWTH OF WIRELESS
19         TELEPHONY?



            109
                     Students residing in college dorms generally receive telephone service from the college
                     or university, not the telephone company. Off-campus housing arrangements, especially
                     with “roommates” would not necessarily have each individual having their own phone
                     line.

                                                      103
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   A124: While it is the case that the number of SBC Ohio access lines has fallen, the overall number of

 2           “wireline” telephones in SBC Ohio territory has continued to grow. Thus, the reduction in SBC

 3           Ohio access lines is better explained by wireline CLEC activities. Considering all wireline access

 4           lines (those provided by CLECs and those provided by SBC), the overall number of wirelines in

 5           SBC Ohio territory has increased by about 3.5% during the 2002-2003 period.110 From June

 6           2002 to June 2003 the number of wireless access lines in Ohio grew by about 772,000 subscribers

 7           or 15.7%111 If substitution of wireless telephones for wireline telephones is occurring, it is not

 8           evident from the continued growth in wireline telephone deployment. Given the magnitude of

 9           wireless subscription growth, it is difficult to comprehend why the number of wireline phones

10           continues to grow, unless most subscribers do not view the wireless service as substitute for their

11           wireline phone.




12   Q125: ARE THERE CULTURAL BARRIERS TO REPLACING A WIRELINE TELEPHONE WITH
13         WIRELESS TELEPHONES?

14   A125: Yes. First, wireless telephone plans bill for usage for both incoming and outgoing calls. High-end

15           wireless plans may offer “buckets” of minutes that can be used at any time, however, exceeding

16           one’s limit may result in charges as much as $0.45 per minute. Consumer aversion to measured



             110
                     This estimate includes DSL lines, which provide a substitute for residential second lines.
                     Growth estimate based on responses to OCC Interrogatories 1 through 7, and 10, as
                     well as data contained in the SBC report “The Status of Competition in Ohio” 4th
                     Quarter 2003.
             111
                    “Telephone Competition: Status as of December June 30, 2003.” Industry Analysis
                    and Technology Division, Wireline Competition Bureau, Federal Communications
                    Commission. June 2003.
     http://www.fcc.gov/Bureaus/Common_Carrier/Reports/FCC-State_Link/IAD/lcom603.pdf

                                                        104
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          local calling is one barrier to the outright replacement of a wireline telephone with a wireless phone.

 2          Second, wireless telephones do not offer a reasonable means to access the Internet. This point is

 3          discussed in more detail below. Finally, for a family to replace a wireline telephone with a wireless

 4          alternative, the transition would require multiple wireless telephones. This would replace the

 5          current single main number for reaching a residence with multiple numbers.                  Even with

 6          wireless/wireline number portability, a main household number would require maintenance of a

 7          specific wireless phone for that purpose.




 8   Q126: ARE WIRELESS TELEPHONES CAPABLE OF PROVIDING DIAL-UP INTERNET
 9         ACCESS?

10   A126: Wireless dial-up is an emerging service, so most consumers would not have the opportunity to use

11          the alternative. To utilize the dial-up wireless service, a special card must be purchased for the

12          user’s PC. Users also pay a premium of up to $80 per month on top of their regular wireless

13          charges.112 Sprint and AT&T market their product as a business service, with business-oriented

14          price points of $50 to $80 per month.




15          Some wireless companies are offering telephones which are capable of providing some basic

16          Internet-related functions, like e-mail and web browsing, however, the “Internet” that is provided

17          through these plans is not the open Internet that is typically available from a dial-up ISP, and is

18          likely a poor substitute for the Internet services available over a dial-up connection. Furthermore,

19          these “wireless Internet” plans impose additional fees, including charges for downloads, based on


            112
                    http://slate.msn.com/id/2087308/


                                                         105
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1           the number of kilobytes transferred.113 Thus, a major factor when considering the decision to

 2           abandon a wireline phone is what to do about Internet access. Broadband is an alternative, but

 3           would likely add another $30 to $40 per month to the cost of replacing the dial-up Internet access

 4           potential associated with a wireline phone.




 5   Q127: CAN A CONSUMER USE A WIRELESS TELEPHONE TO SEND A FAX?

 6   A127: No, the inability to send a fax using a wireless phone is another limitation of the technology and

 7           presents another reason why some consumers would not find it to be a substitute for a wireline

 8           telephone.




 9   Q128: ARE THE ERGONOMICS OF WIRELESS PHONES CONDUCIVE TO VIABLE
10         SUBSTITUTION FOR ALL SEGMENTS OF THE POPULATION?

11   A128: No. Wireless telephones are difficult to hold compared to larger, more ergonomically designed

12           telephone sets available for wireline networks. In addition, wireless handsets present keypads

13           which are often more difficult to see and use. These factors may be highly significant for portions

14           of the population, such as the elderly, or those with physical disabilities.




15   Q129: ARE WIRELINE AND WIRELESS TELEPHONE COMPANIES FINANCIALLY
16         INTERTWINED AND DOES THIS HAVE AN IMPACT ON THE POTENTIAL FOR
17         WIRELESS SUBSTITUTION FOR WIRELINE?

18   A129: Yes. The financial interdependency of wireless and wireline carriers is quite common, with major

19           wireline carriers like SBC, Qwest, BellSouth, Verizon, and Sprint all having major wireless


             113
                     See, for example, information provided by AT&T Wireless at:
                     http://www.attwireless.com/business/plans/mobileinternet/

                                                         106
                                                                  Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                             On Behalf of OCC
                                                                                  Case No. 02-1280-TP-UNC
                                                                                                Public Version


 1   holdings. SBC Ohio’s parent, SBC, is a co-owner of Cingular Wireless, the nation’s second

 2   largest wireless provider. Cingular is in the process of acquiring AT&T wireless, which will

 3   significantly expand Cingular’s reach in Ohio. Verizon is the nation’s largest wireless provider.

 4   Some analysts are pointing to the necessity of combining wireless and wireline business strategies

 5   for success, leaving a company like SBC Ohio in a superior position to wireline providers without

 6   a wireless affiliation, and wireless providers without a wireline affiliation:

 7           “In the longer term only service providers with business models that offer consumers a full
 8           range of both wireline and wireless services, and easy transition between the two
 9           technologies, will succeed,” said Alex Winogradoff, research vice president for Gartner.
10           “Unaffiliated wireless operators will find it difficult to compete against affiliated operators
11           and will likely go out of business unless they adopt a more comprehensive business model
12           or partner with, acquire or be acquired by a wireline operator.”114

13   If this theory is correct, difficult times may lie ahead for wireline CLECs who do not have a wireless

14   affiliate. Some recent examples of the type of revenue increasing wireline/wireless integration

15   strategies available to wireline carriers which also have wireless operations include SBC-Cingular’s

16   MinuteShareK offering, which provides a shared buckets of wireline long distance and wireless

17   minutes.115 SBC-Cingular has also recently introduced a product designed to integrate their

18   wireless customers’ usage with their local telephone service. The new service offering, called

19   FastForward™, is described the SBC as follows:

20           This unique device — designed as a cradle to hold a wireless phone — simply plugs into
21           an electrical outlet. When the Cingular Wireless phone is “cradled,” calls to the wireless
22           phone are forwarded to a designated landline phone, while the wireless phone’s battery
23           is automatically re-charged. Cingular customers with a FastForward device can get
24           unlimited incoming wireless calls (minutes) forwarded to their landline phone in the local
25           calling area — without the minutes counting against their monthly wireless calling plan for


     114
             http://www.cellular-news.com/story/9606.shtml
     115
             http://www.bizjournals.com/sanantonio/stories/2003/06/02/daily28.html

                                                  107
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1                   just $2.99 per month plus the cost of the device. The service is free to SBC residential
 2                   local phone company customers who receive a single bill for Cingular wireless and landline
 3                   services, and BellSouth customers who sign up for a combined bill and two other
 4                   features.116

 5           The FastForward™ device clearly illustrates the type of relationship firms like SBC are cultivating

 6           with their customers who use wireless, one in which their wireless phone is used in conjunction with

 7           the wireline phone—not as a substitute for a wireline phone. SBC also has the ability to market

 8           wireless and wireline service bundles through integrated ordering systems which allow customer

 9           representatives to jointly market wireline and wireless services. Attachment TRR-3 provides SBC

10           documents relating to SBC’s plans with regard to the process of wireline/wireless integration.




11   Q130: HOW DO YOU SEE WIRELINE/WIRELESS INTEGRATION AFFECTING THE
12         ECONOMIC VALUE OF SBC OHIO WIRELINE NETWORK?

13   A130: I believe that the integration of wireline and wireless services will increase the economic value of

14           the wireline network. The synergy across the two services will result in increased value to the

15           customer, which will translate into an improved market position for SBC and SBC Ohio. SBC’s

16           position in the wireless market gives the company a competitive advantage over rivals that do not

17           have wireless operations. Cable television companies and major CLECs such as AT&T and MCI

18           do not have a significant wireless presence. In fact, SBC now has ownership of AT&T’s former

19           wireless assets, leaving AT&T to scramble to arrange a wireless resale arrangement with Sprint

20           to gain access to wireless resources.117


             116
                     http://www.sbc.com/press_room/1,,1783,00.html
             117
                     See, “AT&T Will Offer Cellphone Service with Sprint Deal,” Wall Street Journal,
                     May 18, 2004.
                                                                                            (continued...)

                                                        108
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1   Q131: DO COMPANIES WITH COMBINED WIRELESS AND WIRELINE SERVICE
 2         OFFERINGS GENERALLY ENCOURAGE, THROUGH THEIR MARKETING EFFORTS,
 3         WIRELINE CUSTOMERS TO REPLACE THEIR WIRELINE SERVICES WITH
 4         WIRELESS SERVICES?

 5   A131: No. These companies have a vested interest in their subscribers’ continued use of wireline services

 6           and are developing strategies to integrate their customers’ use of wireline and wireless technologies.

 7           Thus, these companies are very careful to avoid marketing messages that might lead customers to

 8           think about abandoning their wireline connection. In addition, carriers who do not have substantial

 9           wireline customer bases in the local market, such as AT&T, Nextel, and T-Mobile, do not focus

10           on wireless displacement messages.118 Long distance carriers such as AT&T and Sprint, who also

11           have substantial long distance business operations based on wireline access networks, face a risk

12           of cannibalizing wireline long distance revenues and thus must also tread lightly with regard to

13           wireline displacement messages.




14   Q132: DOES WIRELESS SERVICE QUALITY INHIBIT A WIRELESS TELEPHONE’S
15         DESIRABILITY AS A REPLACEMENT FOR A WIRELINE TELEPHONE?

16   A132: Yes. Wireless coverage areas do not provide uniform signal strength and may have “dead zones,”

17           where either no signal or a very weak signal can be received. If your home is in a dead zone, a

18           wireless telephone may not be an alternative to wireline. Furthermore, receiving a wireless signal

19           within a building may be difficult even when a signal is available out-of-doors. Thus, a wireless

20           phone may not provide a very good alternative to a wireline phone when walking from room to


             117
                (...continued)
                      http://online.wsj.com/article_print/0,,SB108482917099713776,00.html
             118
                     “U.S. Wireless Displacement of Wireline Access Lines Forecast and Analysis, 2003-
                     2007,” p. 11. Scott Ellison, IDC Analyze the Future.

                                                         109
                                                              Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                         On Behalf of OCC
                                                                              Case No. 02-1280-TP-UNC
                                                                                            Public Version


 1   room (or into the basement) in a home or apartment. Even where signal strength is adequate to

 2   use the phone indoors, wireless telecommunications providers have incentives to oversubscribe

 3   their services because there are no checks on service quality. The wireless industry practice of

 4   adding customers without expanding network capacity may improve revenues, but it places

 5   increased demands on wireless carrier infrastructure which negatively affects consumers. During

 6   peak calling periods overwhelmed wireless networks have difficulty meeting user demand, leading

 7   to blocked and dropped calls.119 During the August 2003 blackout that affected the Eastern U.S.,

 8   cellular systems did not perform well due to insufficient network capacity and battery back-ups at

 9   cell towers that only provided a few hours of auxiliary power.120 A 2003 survey of wireless users

10   conducted by Consumer Reports magazine indicates that in the week prior to the survey 10

11   percent of respondents experienced conditions where they could not get service, 14 percent

12   experienced dropped calls, and 11 percent experienced poor call quality.121 Another survey

13   conducted on behalf of the U.S. General Accounting Office (GAO) also indicates service quality

14   problems. GAO estimates that 22 percent of wireless users were unable to successfully complete

15   10 percent or more of their calls.122 Consumers are still charged for calls that are dropped by the


     119
               See, for example, “Calls in Question: Why are wireless networks still plagued by
               dropped signals, bad reception and ‘dead zones’?” Wall Street Journal, September
               23, 2002. See also, Letter from Senator Charles E. Schumer to FCC Chairman
               Michael Powell, November 24, 2002. Available at:
           http://schumer.senate.gov/SchumerWebsite/pressroom/press_releases/PR01333.html

     120
              “Blackout gives cell phones a black eye,” CNET News.com. August 15, 2003.
              http://news.com.com/2100-1039-5064689.html

     121
              “Three steps to better cellular,” Consumer Reports, February 2003, p. 16.

     122
              “FCC Should Include Call Quality in Its Annual Report on Competition in Mobile
                                                                                      (continued...)

                                                110
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1           service provider, or may even be charged for not answering their calls quickly enough.123 These

 2           service quality issues make wireless telephones an inferior product when compared to wireline

 3           phones.124


 4   Q133: DO WIRELESS COMPANY MARKETING PRACTICES PROVIDE A BARRIER TO
 5         CUSTOMER EXPERIMENTATION WITH DROPPING WIRELINE SERVICE WITH A
 6         WIRELINE REPLACEMENT?

 7   A133: As is discussed further below, to replace a wireline phone with a wireless offering will require the

 8           purchase of a high-end wireless calling plan. Wireless calling plans require long-term contracts,

 9           rather than the month-to-month services associated with a wireline telephone. These long-term

10           wireless service contracts have high fees for early termination. Long-term contracts with early

11           termination fees present a barrier for residential subscribers who might consider trying to switch

12           away from wireline service. If the wireless replacement did not work out, and wireline service

13           needed to be restored, not only would the consumer face the prospect of a service establishment

14           charge from the wireline provider, but a hefty service termination fee from the wireless provider.

15           As some wireless carriers impose termination fees as high as $200 per number associated with

16           a wireless account, the prospect of experimentation with abandoning wireline becomes highly




             122
                (...continued)
                      Phone Services.” General Accounting Office April 2003, p. 3. Available at:
                      http://www.gao.gov/new.items/d03501.pdf

             123
                     AT&T Wireless recently began charging for calls that go unanswered for more than 30
                     seconds. See, “Fees Hidden in Plain Sight, Companies Add to Bottom Line.” New
                     York Times, December 28, 2002.
             124
                     See also: “U.S. Wireless Displacement of Wireline Access Lines Forecast and
                     Analysis, 2003-2007,” p. 10. Scott Ellison, IDC Analyze the Future.

                                                       111
                                                                        Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                   On Behalf of OCC
                                                                                        Case No. 02-1280-TP-UNC
                                                                                                      Public Version


 1          risky.125 As is discussed elsewhere, replacing wireline with wireless would likely require the

 2          purchase of multiple phones and thus subscribers could face very high termination fees.




 3   Q134: IF SBC OHIO WERE TO CONTINUOUSLY RAISE LOCAL EXCHANGE SERVICE
 4         PRICES, WOULDN’T ITS CUSTOMERS EVENTUALLY FIND WIRELESS TO BE A
 5         VIABLE ALTERNATIVE?

 6   A134: Focusing simply on the price point, and ignoring the service quality and cultural barriers to

 7          switching, it is possible that some customers might. It is important to evaluate the price point

 8          associated with a wireless plan sufficient to substitute for wireline service. For a recent period, the

 9          average number of local minutes associated with an SBC Ohio access line was about 1,650 per

10          month. To evaluate issues involved in switching from a wireline phone to an entirely wireless

11          household, I utilized this usage level to help identify the type of calling plan needed.




12          Given that some portion of the average minutes per line for SBC Ohio is associated with dial-up

13          Internet connectivity, the 1,650 minutes per month should be reduced to generate an estimate of

14          the number of voice minutes per month. Based on information regarding Internet usage, I would

15          estimate that the 1,650 minutes per month average would include, on average, approximately 150

16          minutes of Internet dial-up time.126 This would indicate that approximately 1,500 minutes per


            125
                    T-Mobile imposes a $200 per number early termination fee, Cingular imposes a $150
                    per number early termination fee, Verizon imposes a $175 per number early termination
                    fee. (Information obtained from service terms at carrier web sites.)

            126
                    This estimate is based on information regarding frequency of Internet access associated
                    with a study conducted by the Pew Foundation. See:
                    http://www.pewinternet.org/reports/toc.asp?Report=98

                                                                                                       (continued...)

                                                         112
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1       month reflect average voice usage. Recognizing the wireless plans offer “peak” and “off-peak”

 2       distinctions,127 I conservatively selected plans offering in the 950-1,200 “anytime minutes” range

 3       to identify price points for wireless plans.128 Reviewing various wireless plans indicates that current

 4       prices for wireless service providing this range of anytime minutes range from $60 to $90 per

 5       month, not including taxes, surcharges and fees. I believe this price range is conservative. The high

 6       minute of use charges associated with overrunning the “bucket” of “anytime minutes” associated

 7       with a wireless plan would provide an incentive to purchase plans with more “anytime” minutes as

 8       insurance against costly usage overruns.




 9       The $60 to $90 per month charge includes one telephone. Additional charges will apply to add

10       telephone lines. It is important to note that as wireless phones are “mobile,” it would be likely that

11       replacing a wireline connection with wireless would require multiple units for a typical family. If the

12       sole wireless phone “walks away” with an individual family member, then risk and inconvenience

13       are imposed on the remaining family members left without a telephone.

14   Q135: FOR A TYPICAL FAMILY, WHAT PRICE POINT DO YOU ESTIMATE FOR
15         REPLACING A WIRELINE PHONE WITH A WIRELESS ALTERNATIVE?



         126
            (...continued)
                  The assumptions based on the Pew study are that approximately 50% of households
                  have dial-up Internet access, 55% of households go online every day, and that average
                  time online is 15 minutes.

         127
                 Night and weekend minutes usually apply for a relatively small portion of the week.
                 Night minutes typically run from 9 p.m. to 6 a.m. The remainder of the day must be
                 covered by the “anytime minutes,” if high minute-of-use charges are not to apply.

         128
                 Variation in the service offerings by wireless providers did not allow for evaluation of a
                 fixed number of anytime minutes.

                                                      113
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   A135: Replacing a wireline phone with a wireless phone would most likely mean purchasing a relatively

 2          “high end” wireless plan, and the addition of multiple handsets could easily lead the price per

 3          household of “basic wireless service” to climb over $100 per month, not even addressing the

 4          Internet access issue. Of course these high-end wireless plans include long distance and may also

 5          include other features, like call waiting and voice mail, thus specifics of consumer preferences and

 6          usage would influence the price-point at which rising wireline prices would lead subscribers to

 7          abandon the wireline phone. However, for individuals who do not use much long distance, or who

 8          value their dial-up connection to the Internet, considerable price increases could be imposed by

 9          SBC Ohio without the wireless alternative providing a constraint, not even considering factors such

10          as lower service quality. Thus, there is a highly plausible explanation for the observed outcome

11          of wireless subscription growing very rapidly in Ohio, while wirelines in SBC Ohio territory

12          continue to grow—the overwhelming majority consumers do not view wireless as a substitute for

13          wireline.




14   Q136: YOU INDICATE THAT IT IS POSSIBLE THAT SOME CUSTOMERS MIGHT FIND
15         THE WIRELESS ALTERNATIVE ATTRACTIVE, BECAUSE THEY ALSO USE
16         SIGNIFICANT AMOUNTS OF LONG DISTANCE CALLING AND USE ADVANCED
17         SERVICES. DO MARKET ANALYSTS AGREE ON THE MAXIMUM NUMBER OF
18         SUBSCRIBERS WHO MIGHT MAKE THE SWITCH?

19   A136: There appears to be some consensus on this issue. Both the Gartner Group and IDC point to an

20          eventual displacement of about 10% of wireline phones by wireless. However, the horizon for this

21          transition stretches as far as 2007. Thus, in the long run, 90% of the market is viewed as not likely




                                                        114
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1          to consider abandoning their wireline connection for wireless.129




 2   Q137: YOUR DISCUSSION OF WIRELESS FOCUSES ON THE RESIDENTIAL MARKET
 3         SEGMENT. IS THERE A GREATER POTENTIAL FOR WIRELESS SUBSTITUTION ON
 4         THE BUSINESS SIDE?

 5   A137: No, there is a much smaller potential for business migration to wireless. Businesses utilize PBXs

 6          and corporate voice-mail, making wireless a less attractive substitute because wireless plans do

 7          not offer the high degree of customization that is available from these systems. Furthermore,

 8          business migration to voice over IP, exploiting the economics from integrating internal voice and

 9          data networks does not have a wireless component. Businesses certainly make use of wireless

10          telephony, but as an extension to their wireline systems. Displacement of business wireline

11          networks with wireless is not expected anytime soon. 130




12   Cable Companies and Voice Over IP

13   Q138: DO YOU AGREE WITH DR. VANSTON THAT THE THREAT POSED BY CABLE
14         COMPANY VoIP WARRANTS A REDUCTION IN DEPRECIATION LIVES?

15   A138: No. Cable telephony is still in its infancy, with major cable companies such as Cox, Comcast, and

16          Time Warner essentially in the trial phase of the rollout. Thus, it is highly speculative to claim, as

17          Dr. Vanston does, that significant numbers of subscribers will be switching to cable telephony in




            129
                    See: “U.S. Wireless Displacement of Wireline Access Lines Forecast and Analysis,
                    2003-2007,” p. 11. Scott Ellison, IDC Analyze the Future. For Gartner Group
                    projection, see, http://www.cellular-news.com/story/9606.shtml
            130
                    “U.S. Wireless Displacement of Wireline Access Lines Forecast and Analysis, 2003-
                    2007,” Scott Ellison, IDC Analyze the Future.

                                                        115
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1          the near future.131 It is notable that cable companies are marketing cable telephone service as part

 2          of service bundles, which typically target high-volume toll users. The recent Illinois proposed

 3          decision in SBC Illinois’ UNE pricing case rejected similar claims regarding cable’s potential

 4          impact:

 5                    Moreover, SBC’s competition forecasts are extreme. SBC argues that cable telephone
 6                    could achieve 25% or more market share within a few short years. Staff counters that if
 7                    recent growth trends in cable telephony are maintained then it would take until the year
 8                    2045 for cable telephone to reach about 9% of the market.132


 9          The Commission should not accept Dr. Vanston’s claims regarding the potential for cable

10          telephony as providing a basis to reduce SBC Ohio’s depreciation lives.




11   The Impact of Competition on the Economic Value of SBC Facilities

12   Q139: IF FACILITIES-BASED COMPETITION EMERGES IN SBC OHIO’S SERVICE
13         TERRITORY, DO THE ACCESS LINES WHICH SBC OHIO LOSES BECOME
14         “STRANDED,” AS DR. VANSTON SUGGESTS?

15   A139: As I have discussed earlier in this testimony, evidence of facilities-based competition in SBC

16          Ohio’s service area is very limited, especially for residential subscribers. If facilities-based

17          competition were to emerge, and SBC Ohio were to begin to lose customers to the alternative

18          technology, there is no question that there would be some impact on the economic value of the

19          facilities, however, the impact would not be as Dr. Vanston suggests. Dr. Vanston indicates that

20          facilities-based competition will “strand” investment, “a lost access line loses its value when the




            131
                      Dr. Vanston, direct testimony, pp. 34-35.
            132
                      Illinois Proposed Decision, p. 73.


                                                       116
                                                                 Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                            On Behalf of OCC
                                                                                 Case No. 02-1280-TP-UNC
                                                                                               Public Version


 1   customer switches to a facilities-based competitor.”133 However, this reasoning is highly unrealistic

 2   and ignores the dynamic of facilities-based competition.




 3   Dr. Vanston assumes that once an SBC Ohio customer is “lost” to a competitor, that the loss is

 4   “forever.” There are a number of factors which are likely to make this an unreasonable assumption.

 5   If SBC Ohio loses a customer to a rival, SBC Ohio’s network facilities continue to provide SBC

 6   Ohio the opportunity do business with the customer. In economic terms, the value changes from

 7   that directly supplied by a business relationship with the customer, which was the situation prior to

 8   the customer loss, to an option value. In other words, the network facility continues to provide

 9   value due to the ability to serve the customer (or some other customer) in the future. SBC Ohio’s

10   marketing strategies include the use of “WinBack” programs. Where a customer has been lost to

11   facilities-based entry, the ability to win the customer back will depend on the continuing availability

12   of SBC Ohio facilities. Likewise, at any specific location (either residential or business), there is

13   customer churn which is entirely unrelated to choice of telecommunications provider. For example,

14   residential and business subscribers change locations, which then would create a competitive

15   opportunity to win the business of the subscriber in the location where the churn has occurred.

16   Absent the availability of facilities, the opportunity to gain the subscribers’ business is diminished.

17   Thus, when considering the “stranded investment” arguments offered by telephone company

18   representatives, it is appropriate to note that while the emergence of facilities-based competition

19   has the potential to reduce economic value of assets, it does not necessarily drive the value of those

20   assets to zero, as is suggested by Dr. Vanston.


     133
             Dr. Vanston, direct testimony, p. 42, line 20.

                                                  117
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q140: ARE YOU SAYING THAT THERE WILL BE NO CHANGE IN THE ECONOMIC
 2         VALUE OF SBC OHIO’S ASSETS AS A RESULT OF FACILITIES-BASED
 3         COMPETITION?

 4   A140: No. However, my point is that the impact on the value of the facilities is not as dramatic as is

 5          suggested by Dr. Vanston’s “stranding” argument.




 6   Q141: PLEASE SUMMARIZE YOUR DISCUSSION OF THE IMPACT OF COMPETITION
 7         ON THE ECONOMIC VALUE OF SBC OHIO’S FACILITIES.

 8   A141: There is strong evidence that the type of competition which SBC Ohio will face in the future will

 9          continue to rely on the use of SBC Ohio facilities. As a result, the impact on economic value of

10          these facilities is not as dramatic as has been suggested by Dr. Vanston. In addition, countervailing

11          market forces, such as the strategic use of resale, wireless complementarity, and the growth of DSL

12          services will enhance the economic value of SBC Ohio’s facilities. Furthermore, the loss of an

13          access line to a facilities-based competitor, a situation which has not been the norm to date in SBC

14          Ohio’s service territory (especially in the residential market segment), does not drive the economic

15          value of SBC Ohio facilities to zero, as the “stranded” investment argument offered by Dr. Vanston

16          suggests.




17   Technological Change and Economic Value—The Impact of Broadband

18   Q142: DO YOU AGREE WITH DR. VANSTON THAT A VARIETY OF SBC OHIO FACILITIES
19         WILL BE OBSOLETE WITHIN FIVE YEARS DUE TO THE NEED TO TRANSITION
20         TO ACCESS BANDWIDTHS OF 24 MBPS AND HIGHER?

21   A142: No, I believe that Dr. Vanston’s scenario is highly speculative. Ultra-highspeed broadband to the

22          home presents a highly risky investment strategy, whether pursued with fiber to the home, or some



                                                        118
                                                               Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                          On Behalf of OCC
                                                                               Case No. 02-1280-TP-UNC
                                                                                             Public Version


 1   other strategy which pushes fiber very close to the customer premise. I believe that the reluctance

 2   of firms like SBC to move fiber optic cable into distribution plant reflects market realities which

 3   make the demand for the extremely high bandwidth associated with fiber doubtful in the near term.




 4   Before continuing with this discussion, it is appropriate to reveal my personal bias, I believe that

 5   broadband access is a critical technology. I have broadband Internet access at home and at my

 6   office. It is indispensable for me. However, my opinion does not appear to reflect the opinion of

 7   even the majority of Internet users in the U.S. About 47% of residential Internet users are

 8   purchasing broadband, which translates to something less than 25% of all households. In addition,

 9   information gathered by the Neilsen/Netratings report that as of April 2004, most users in the US

10   connect to the Internet using dial-up modems of 56Kbps or less: 42.74% use 56Kbps modems,

11   6.99% use 28/33.3Kbps, and 2.4% use 14.4Kbps modems. In total, 52.13% of home users in

12   the US connect to the Internet at 56Kbps or less.134




13   These statistics should justifiably make any business considering launching ultra-high-speed

14   broadband to the home initiative wary. The fact that nearly 10% of Internet users still choose to

15   rely on modem speeds of 33.6kbps and below is significant and reflects several facts about the

16   “broadband” revolution which must be taken into account when considering the impact of

17   technological change on the continued viability of dial-up and DSL and thus copper distribution

18   cable.



     134
              The data is based on a Neilsen survey and reported at:
              http://www.websiteoptimization.com/bw/0403/

                                                119
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1   Q143: WHY HAVE RESIDENTIAL INTERNET USERS BEEN SLOW TO ADOPT BROADBAND?

 2   A143: Price is likely a contributing factor. However, there are technical considerations as well. The

 3          Internet delivers relatively good performance for popular applications with relatively little access

 4          bandwidth required. For example, e-mail, instant messaging, and chat, all very popular

 5          applications, work very well over a dial-up connection. Web browsing over a dial-up connection

 6          can be less satisfying, however, the developers of web sites are acutely aware of the fact that their

 7          audience is most likely to be using dial-up technology, thus web-site-design is driven by the need

 8          to deliver reasonable performance over dial-up facilities, meaning that web content may be

 9          “dumbed down” to perform better over dial-up connections. In addition, web developers can take

10          advantage of processing power on the user’s personal computer to serve as a substitute for more

11          bandwidth—for example, use of Macromedia’s “Flash” and “ShockWave” programs enable

12          dynamic web content and good quality streaming animation over dial-up connections due to the

13          innovative design of the playback software that the user loads on a computer. The higher

14          bandwidth that is available over digital subscriber lines enables excellent performance with other

15          Internet applications, such as web browsing, streaming media, and large file transfer. However,

16          given current broadband Internet access prices, consumers have not entirely “jumped ship” from

17          dial-up technology. This is in large part because of the performance of popular Internet

18          applications which easily scale to available bandwidth and provide adequate service quality at

19          transmission speeds associated with the dial-up option.




20   Q144: WHAT TECHNOLOGY WOULD JUSTIFY THE VERY HIGH BIT-RATES PROJECTED
21         BY DR. VANSTON TO BE “JUST AROUND THE CORNER”?



                                                        120
                                                                          Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                     On Behalf of OCC
                                                                                          Case No. 02-1280-TP-UNC
                                                                                                        Public Version


 1   A144: To justify consumer demand for the access data rates which Dr. Vanston argues is at hand (i.e.,

 2          24 Mbps and above), the only content that can warrant broadband speeds of this caliber is video,

 3          especially high-definition video. Justifying the replacement of existing access facilities for the

 4          delivery of video services over a shared IP network platform over fiber presents a risky business

 5          proposition for a number of reasons.




 6          Those interested in deploying the type of high-end broadband networks envisioned by Dr. Vanston

 7          face two fundamental market problems: (1) there are already multiple competing video delivery

 8          networks (cable TV and direct broadcast satellite) and (2) that video storage (DVD and

 9          videocassette) is a substitute for video provided over a fiber-based data transmission platform.

10          Multiple video delivery networks ensure that the delivery of video will face strong competition if

11          another delivery path (e.g., fiber to the home) is pursued. Furthermore, DVDs offer a number

12          of other advantages that video delivered over networks cannot, for example, the ability to view the

13          content repeatedly, and to share DVDs with friends and relatives.




14          To summarize on this point, Dr. Vanston’s view does not take into account the factors that I have

15          discussed above, i.e., the fact that most Internet users today still dial-up; the ability of Internet

16          applications to economize on the demand for bandwidth; and the fact that existing broadband

17          alternatives, including SBC Ohio’s own DSL service, provide a very good user experience. In

18          fact, SBC Ohio has experienced very strong demand growth for its copper-based DSL service,

19          and this success will extend the life of copper facilities.




                                                         121
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1   Q145: HAS SBC ADOPTED DR. VANSTON’S VIEWPOINT WITH REGARD TO FIBER
 2         DEPLOYMENT?

 3   A145: No. Perhaps recognizing the risks associated with building a high-capacity broadband access

 4          network to deliver video services, SBC has elected to form an alliance with Dish Networks to

 5          deliver video using satellite facilities. SBC CEO Edward Whitacre is not optimistic regarding fiber

 6          to the premises (FTTP):

 7                  SBC Chief Edward Whitacre Jr says he just “doesn’t see the incentive for FTTP” (fiber
 8                  to the premises). Despite the fact that the new FCC triennial review frees the Bells from
 9                  having to share new fiber deployments with competitors, the SBC boss, speaking at a
10                  Morgan Stanley conference, said the technology was still too cost prohibitive to be
11                  seriously considered. Whitacre pegged FTTH costs at $2,000 per home, and notes he
12                  probably wouldn’t be any more likely to consider FTTH even if prices dropped several
13                  hundred dollars.135

14          An analyst with Telecom Pragmatics, Inc. notes “SBC will likely wait and see how Verizon does.

15          . . FTTP will happen, but its going to be slow and they will only do the easy jobs first. “136 Even

16          representatives of Verizon, which is viewed as the industry leader in the fiber to the home vision,

17          indicate that DSL is likely to be the dominant technology for a number of years and that Verizon

18          continues to upgrade its network to provide DSL. 137




19   Q146: PLEASE SUMMARIZE YOUR DISCUSSION OF TECHNOLOGICAL SUBSTITUTION
20         AND ITS IMPACT ON THE ECONOMIC VALUE OF SBC OHIO’S FACILITIES.

21   A146: Dr. Vanston’s view of technological substitution presumes that copper distribution cables will be



            135
                    “‘No Incentive’ For Fiber: SBC chief not optimistic about FTTP,” Broadband
                    Report.com, 9-10-03, http://www.broadbandreports.com/shownews/32630
            136
                    “”SBC Ratchets up PON Politics,” Light Reading–Networking the Telcom Industry,
                    September 11, 2003, http://www.lightreading.com/document.asp?doc_id=40031
            137
                    See, “Verizon Exec Talks up FTTP,” Network World, 6/9/03.
                    http://www.nwfusion.com/news/2003/0609lacouture.html

                                                       122
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


 1          eliminated from use in a relatively short period of time. He assumes that consumers will have ever-

 2          growing demand for bandwidth and will be willing to pay for the privilege. I disagree with Dr.

 3          Vanston’s assessment. I believe that a much more realistic scenario is that the transmission speeds

 4          associated with current broadband technologies, such as that provided by existing DSL services,

 5          will provide all the bandwidth consumers need for a considerable length of time. Fiber to the home

 6          will likely emerge gradually, with high-end customers in new affluent subdivisions seeing the

 7          deployment of the technology first. The rapid obsolescence of copper distribution cable argued

 8          by Dr. Vanston is simply not supported by market reality.




 9   Q147: HOW HAVE OTHER COMMISSIONS RESPONDED TO DR. VANSTON’S CLAIMS?

10   A147: Recent decisions have rejected Dr. Vanston’s position. For example, the Illinois proposed

11          decision rejected the SBC Illinois proposal to use financial reporting lives and noted:

12                  We find it unrealistic that the circuit-switched, copper network will begin to become
13                  obsolete in five years as suggested by SBC witness Vanston. SBC has deployed
14                  substantial quantities of DSL lines, adding 377,000 in the fourth quarter of 2003. . . . We
15                  agree with AT&T/MCI that this serves to increase the life and usefulness of copper plant.


16                  Further, the Commission agrees with Staff that Vanston’s forecast of the growth of
17                  broadband customers is excessive. Vanston forecasts penetration of 90% in 2020 that he
18                  argues requires very high levels of availability, i.e., at least 95%. . . . Staff notes that
19                  currently, however, cable companies serve two-thirds of broadband subscribers. We are
20                  concerned that under SBC’s proposal, ratepayers who don’t subscribe to broadband, or
21                  who obtain broadband from the cable companies, would be required to pay higher
22                  telephone rates just to better position the telephone companies to compete in the
23                  broadband market.

24                  Moreover, SBC’s competition forecasts are extreme. SBC argues that cable telephone
25                  could achieve 25% or more market share within a few short years. Staff counters that if
26                  recent growth trends in cable telephony are maintained then it would take until the year
27                  2045 for cable telephone to reach about 9% of the market. Similarly, SBC argues that


                                                       123
                                                                      Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                 On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                                    Public Version


 1                  competition from wireless will replace wireline communication completely in 20% of
 2                  households by 2005. Staff points out, however, that to date only 2% of households rely
 3                  exclusively on wireless and, to reach SBC’s forecast, would require almost three times as
 4                  many customers to switch per year as have switched to date.138

 5          Proposed Orders emanating from a proceeding before the California Public Utilities Commission

 6          are also skeptical of Dr. Vanston’s claims and reversed the California Commission’s previous use

 7          of financial reporting lives, adopting instead the FCC’s depreciation lives which I recommend be

 8          used for purposes of this proceeding.

 9                  We agree. . . that the widespread deployment of DSL technolgy, which uses copper cable
10                  to provide high-speed Internet service, illustrates that competition can actually lengthenthe
11                  economic lives of some assets. . . . Thus, the FCC’s asset life findings contradict Vanston’s
12                  assumptions regarding rapid substitution of new technologies for existing ones.139


13   The FCC’s Current Depreciation Lives

14   Q148: IF THE FCC’S CURRENT DEPRECIATION LIVES WERE INSUFFICIENT, WHAT
15         OUTCOME WOULD BE EXPECTED?

16   A148: First, it would be expected that depreciation rates would be outstripped by retirement rates.

17          Second, and interrelated to the first expected result, it would also be expected that depreciation

18          reserves would decline. Neither of these expectations is supported by the data available from the

19          FCC’s ARMIS records. Table 7, below, shows that for the last five years, SBC Ohio’s retirement

20          rates have fallen well below the depreciation rates. This indicates that depreciation rates have been

21          set to recover more than the actual mortality



            138
                    Administrative Law Judges’ Proposed Order in case 02-0864, Illinois Bell Telephone
                    Company:       Filing to increase Unbundled Loop and Nonrecurring Rates. May
                    6, 2004, pp. 72-73.

            139
                    Proposed Decision of Judge Duda, p. 114, and Proposed Alternative Decision of
                    Commissioner Wood, p. 111. Application 01-02-024, May 3, 2004.

                                                        124
                                                                                Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                           On Behalf of OCC
                                                                                                Case No. 02-1280-TP-UNC
                                                                                                              Public Version

 1        Table 7: TPIS, Retirements, and Depreciation, SBC Ohio, 1999-2003 (Dollars in thousands)
                    TPIS (BOY)   TPIS (EOY)   Retire-    Accumulated      Depreciation   Retire-     Depreciation   Accumulated
                                              ments      Depreciation      Accruals       ment          Rate        Depreciation
 2        Year                                              (BOY)                         Rate                       as a % of
                                                                                                                       TPIS


                        A            B           C             D               E           C/A        E/(A+B)/2         D/A

 3           1999   $6,552,280   $6,843,618 $165,012         $3,436,252       $415,690     2.52%            6.21%         52.44%
 4           2000   $6,843,616   $7,232,194 $135,484         $3,699,699       $444,970     1.98%            6.32%         54.06%
 5           2001   $7,232,194   $7,734,569  $77,634         $4,022,933       $473,026     1.07%            6.32%         55.63%
 6           2002   $7,734,569   $8,005,141 $149,107         $4,422,270       $509,875     1.93%            6.48%         57.18%
 7           2003   $8,005,142   $8,161,878 $105,459         $4,786,501       $520,285     1.32%            6.44%         59.79%




 8            of equipment. Also, depreciation reserves have grown in that period.


 9   SBC Ohio’s Proposed Depreciation Lives

10   Q149: HOW DOES SBC OHIO UTILIZE DR. VANSTON’S ANALYSIS?

11   A149: SBC Ohio utilizes Dr. Vanston’s views to justify its use of financial lives to be used in the

12           LoopCAT model.




13   Q150: DO YOU AGREE WITH THE USE OF FINANCIAL LIVES?

14   A150: No. Financial reporting lives are determined by the management of SBC. Although these lives are

15           subject to audit by a public accounting firm, the auditors do not attest to the accuracy of the asset

16           lives selected by management. Financial reporting lives, based on Generally Accepted Accounting

17           Principles (GAAP), do not necessarily reflect expected changes in economic value. Rather,

18           financial reporting lives are based on principles of financial conservatism which are associated with

19           the protection of investors. The FCC has addressed the financial conservatism issue in assessing

20           ILEC requests for the latitude to use financial lives in the context of ILEC price cap plans:

21                    ‘GAAP is guided by the conservatism principle which holds, for example, that, when
22                    alternative expense amounts are acceptable, the alternative having the least favorable effect


                                                               125
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                                     Public Version


 1                  on net income should be used.’ The Commission concluded that, although conservatism
 2                  is effective in protecting the interests of investors, it may not always serve the interests of
 3                  ratepayers, and did not offer adequate protection for ratepayers in the case of depreciation
 4                  accounting. We are not persuaded that the role of the conservatism principle in GAAP has
 5                  changed or that we should change our previous decision. 140

 6          The worst-case conservatism of GAAP is not consistent with establishing economic depreciation

 7          rates. While the FCC has now indicated in its Triennial Review Order that discretion should be

 8          left with the state commissions with regard to whether financial lives or regulatory lives should be

 9          employed, the FCC’s Wireline Competition Bureau has recently rejected the use of financial lives

10          and found that the FCC’s approved regulatory lives are consistent with the forward-looking

11          perspective of TELRIC:

12                  We reject Verizon’s argument that FCC regulatory lives are not sufficiently forward-
13                  looking. The Commission has used forward-looking asset lives for some time in its
14                  regulation of incumbent LEC depreciation practices, and the asset lives that we adopt here
15                  are the most recent ones prescribed by the Commission. . . .141

16          Thus, I believe that applying the FCC’s lives and salvage values is a reasonable approach to

17          modeling TELRIC costs for SBC Ohio. Dr. Vanston’s analysis does not support an imminent and

18          dramatic change in the economic value of SBC Ohio’s assets.




19   Q151: PLEASE SUMMARIZE THIS SECTION OF YOUR TESTIMONY.

20   A151: SBC Ohio’s proposal to reduce depreciation lives from the levels previously specified by the



            140
                    In the Matter of 1998 Biennial Regulatory Review—Review of Depreciation
                    Requirements for Incumbent Local Exchange Carriers; United States Telephone
                    Association’s Petition for Forbearance from Depreciation Regulation of Price
                    Cap Local Exchange Carriers, CC Docket 98-137, ASD 98-91. Report and Order
                    in CC Docket 98-137, Memorandum Opinion and Order in ASD 98-91, December
                    30, 1999, FCC 99-397, ¶49, footnotes omitted.
            141
                    Virginia Arbitration Order, ¶115.

                                                        126
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                     Case No. 02-1280-TP-UNC
                                                                                                   Public Version


1         Commission is not supported by Dr. Vanston’s claims regarding the impact of technological and

2         market substitution. Evidence points to the FCC’s depreciation rates as being consistent with

3         forward-looking cost modeling, i.e., it is clear that those rates have resulted in depreciation

4         outstripping actual retirements, thus indicating a more rapid depreciation rate than that justified by

5         the historic mortality of equipment. I recommend that lives and salvage values consistent with the

6         FCC’s be utilized.




7   Q152: DOES THIS CONCLUDE YOUR TESTIMONY AT THIS TIME?

8   A152: Yes.




                                                      127
                   Attachment TRR-1
                   Dr. Roycroft’s Vita
Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                          TREVOR R. ROYCROFT
                                               P.O. Box 5681
                                               9065 Echo Lane
                                              Athens, OH 45701
                                            Voice: (740) 593-8552
                                             Fax: (740) 594-8552


Education

Ph.D., Economics, University of California, Davis, 1989.
M.A., Economics, University of California, Davis, 1986.
B.A., Economics, with honors, California State University, Sacramento, 1984.

Fields of Specialization

Industrial Organization and Regulation
Public Finance
Economic History

Experience

Associate Professor, J. Warren McClure School of Communication Systems Management, Ohio University,
September 1994 to present. Granted tenure, Spring 2000. Conducts graduate and undergraduate courses in
regulatory policy and law, and the economics of the telecommunications industry, as well as general education
courses covering telecommunications technology, markets, policy, and the social impact of communications
technology. Conducts research with a focus on the telecommunications industry. Engages in consulting related to
the telecommunications industry. Provides academic advising to undergraduates within the school and across the
university. Serves on department, college, and university committees.

Interim Director, J. Warren McClure School of Communication Systems Management, Ohio University, July 2000
to June 2002. Responsibilities included: program planning, evaluation, and assessment; recruiting faculty and staff;
managing fiscal resources; administering the School’s curriculum; and establishing and maintaining relationships with
internal and external constituencies of the school.

Chief Economist/Acting Chief Economist/Assistant Chief Economist/ Principal Economist, Indiana Office of Utility
Consumer Counselor, May 1991 to June 1994. Conducted research and prepared testimony, cross examination,
and legal briefs to be presented before the Indiana Utility Regulatory Commission in major cases involving gas,
water, electric, and telecommunications utilities. Prepared analysis and comments to be presented before the
Federal Communications Commission. Advised Director of Utility Analysis and the Utility Consumer Counselor
on policy issues; assisted in formulation of policy. Coordinated technical analysis in major cases. Presented agency
policy positions to outside groups. Supervised Economics and Finance Staff of eight professionals. Reviewed and
provided extensive analysis of Economics and Finance Staff testimony.



                                                         1
Experience (continued)

Visiting Assistant Professor, Kenyon College, September, 1989 to May, 1991. Conducted courses in Introductory
Economics (Macro and Micro), Economics of the Public Sector, Industrial Organization, and Economic
Development in the Third World. Rendered college service on award and hiring committees.

Lecturer, California State University, Sacramento, Fall 1987, academic year 1988. Conducted courses in
Intermediate Microeconomic Theory, Introductory Macroeconomic and Microeconomic Theory.

Teaching Assistant, University of California, Davis, 1985-1988. Assisted the professor in conducting courses in
Introductory Macroeconomic Theory, Introductory Microeconomic Theory, and Public Finance.

Research in Progress

Open Access: The Internet and Other Policy Accidents

“Benchmarking the Costs of Competition: What the FCC’s Synthesis Model Says About the Prospects for
Facilities-Based Local Exchange Competition.”

“Evaluating RBOC Productivity as Competition Emerges.”

Publications

“Empirical Analysis of Entry in the Local Exchange Market: the Case of Pacific Bell.” Contemporary Economic
Policy, forthcoming.

“Internet Access,” in Johnson, D. ed. Encyclopedia of International Media and Communications, Academic
Press, April 2003.

“Internet Subscription in Africa: Policy for a Dual Digital Divide.” (With Siriwan Anantho.) Telecommunications
Policy, Vol. 27, Nos. 1-2, February/March 2003.

“The Impact of State and Federal Alternative Regulation Plans on the RBOCs--a State Level Analysis.” in
Telecommunications for the 21st Century. Special issue of The International Journal of Development
Planning Literature. William Baumol and Victor Beker eds. Vol. 16, Nos. 1 & 2, January and April 2001.

“Trouble Reports as an Indicator of Service Quality: The Influence of Competition, Technology, and Regulation.”
(With Martha Garcia-Murrilo.) Telecommunications Policy, Volume 24, No. 10, November, 2000.




                                                      2
Publications, Continued

“The Telecommunications Act--Law of Unintended Consequences?” Public Utilities Fortnightly, Volume 138,
No. 3, February 1, 2000.

“Alternative Regulation and the Efficiency of Local Exchange Carriers--Evidence from the Ameritech States.”
Telecommunications Policy, Volume 23, No. 6, July, 1999.

“The Billy Goats Gruff. A Fairy Tale for the Third Anniversary of the Telecommunications Act of 1996.”
Info: The Journal of Policy, Regulation and Strategy for Telecommunications, Information and Media,
Volume 1, No. 2, April, 1999.

“A Dynamic Model of Incumbent LEC Response to Entry Under the Terms of the Telecommunications Act of
1996.” Journal of Regulatory Economics, Volume 14, November, 1998.

“Ma Bell’s Legacy: Time for a Second Divestiture?” Public Utilities Fortnightly. Vol 136, No. 12, June 15,
1998.

“The Telecommunications Act of 1996: An Unfunded Mandate for the States.” (With Phyllis Bernt.) Central
Business Review, Volume XV, No. 2, Summer 1996.

Commissioned Reports

“Wireless Consumer Protection: A Model Bill for the States.” AARP Research Center, September, 2003.

“The End of Telecommunications? An Epilogue to Tangled Web: The Internet and Broadband Open Access
Policy.” AARP Research Center, June, 2002.
Available at: http://research.aarp.org/consume/2002_10_tangled_1.html

“Tangled Web: The Internet and Broadband Open Access Policy.” AARP Research Center, January, 2001.
Available at: http://research.aarp.org/consume/d17331_tangled_1.html

Conference Papers

“The Impact of State and Federal Alternative Regulation Plans on the RBOCs--a State Level Analysis,” July 1999.
Presented at the Western Economic Association International Annual Meeting, San Diego, California.

“The Billy Goats Gruff. A Fairy Tale for the Third Anniversary of the Telecommunications Act of 1996,” June,
1999. Presented at the Academic Seminar at the 1999 National Cable Television Association Convention,
Chicago, Illinois.



                                                      3
Conference Papers (Continued)

“Alternative Regulation and the Efficiency of Local Exchange Carriers--Evidence from the Ameritech States.”
November, 1998. Presented at the 68th Annual Conference of the Southern Economic Association, Baltimore,
Maryland.

“A Dynamic Model of Incumbent LEC Response to Entry Under the Terms of the Telecommunications Act of
1996.” July 1998. Presented at the Western Economic Association International Annual Meeting, Lake Tahoe,
Nevada.

“Do We have the Bugs Out of Telephone Deregulation?” April 1998. Presented at the Law and Policy Division
of the Broadcast Education Association, Las Vegas, Nevada.

“The Telecommunications Act of 1996 and Imposed Costs in the Local Exchange Market: A Dynamic Model of
Incumbent Behavior.” September 1997. Presented at the Telecommunications Policy Research Conference,
Arlington Virginia.

“Towards an Advanced Information Infrastructure,” August 1995. Presented to the National Association of
Regulatory Utility Commissions' Annual Regulatory Studies Program at Michigan State University.

“Sorting, Bonding, and Barriers to Entry: Strategies of the Entry Concerned Firm,” July 1990. Presented at the
Western Economic Association Meetings, San Diego, California.

Additional Presentations

“Broadband Open Access.” Presented to AARP’s National Legislative Council. October, 2000. Washington,
D.C.

“Telecommunications Policy, Markets, and Regulation–Who’s On First?” Presented to the Maryland Office of
Peoples’ Counsel and Maryland Public Service Commission. October, 2000. Baltimore, MD.

“Broadband Open Access–Implications for the Internet and Consumers.” November 1999. Panelist at the
National Association of Utility Consumer Advocates Annual Convention. San Antonio, Texas.

“Validation of Proxy Cost Models.” January 1997. Panel discussant at the Federal Communications Commission
workshops on proxy cost models (CC Docket 96-45).

“Impact of the Telecommunications Act of 1996 on Telecommunications Managers.” December 1996. Presented
to members of the Association of Telecommunications Professionals. Columbus Ohio.

“Caveat emptor! Local competition, possible effects on prices and the reality of choice.” October 1995. Presented
at the Public Information Session on Telephone Competition. Dayton, Ohio.

                                                       4
Additional Presentations (Continued)

“Cost Allocation in Network Industries,” August 1995. Presented to the National Association of Regulatory Utility
Commissions' Annual Regulatory Studies Program at Michigan State University.

“Incremental Cost Methodology in Telecommunications,” June 1995. Presented to the Ohio Office of Consumers'
Counsel.

“Regulatory Issues Connected with the Implementation of the Clean Air Act Amendments of 1990,” August 1993.
Presented at the Indiana Bar Association's Utility Law Section Summer Meetings.

“Consumer Perspectives on the Ameritech Customer's First Plan,” August 1993. Presented at the Ameritech
Regional Regulatory Committee Ad Hoc Working Group Meeting.

“Consumer Perspectives on Universal Telecommunications Service,” December 1992. Presented at the Indiana
Utility Regulatory Commission Workshops on Regulatory Flexibility in Telecommunications.

Honors

Competitive paper finalist. The Academic Seminar at the 1999 National Cable Television Association Convention,
Chicago, Illinois. Paper title: “The Billy Goats Gruff. A Fairy Tale for the Third Anniversary of the
Telecommunications Act of 1996.”

Courses Taught

Competition and Market Structure in Network Industries (graduate and undergraduate), Ohio University
Communication Regulatory Policy, Ohio University
Applications of Common Carrier Regulation, Ohio University
Introduction to Common Carrier Regulation, Ohio University
Introduction to Communication Systems Management, Ohio University
Consumer Issues in Communication Systems Management, Ohio University
Topical Seminar (New Technologies and Telecommunication Policy), Ohio University
Topical Seminar (The Telecommunications Act of 1996), Ohio University
Special Studies in Communication Systems Management, Ohio University
Economics of the Public Sector, Kenyon College
Industrial Organization, Kenyon College
Economic Development in the Third World, Kenyon College
Intermediate Microeconomics, California State University, Sacramento
Microeconomic Principles, Kenyon College; California State University, Sacramento
Macroeconomic Principles, Kenyon College; California State University, Sacramento



                                                       5
Service

Faculty Advisor, University College, Ohio University, 1998-2004
Member, Baker Fund Committee, Ohio University, 2003-2006
Member, College of Communication Curriculum Committee, Ohio University, 2003-2004
Chair, College of Communication Dean’s Evaluation Committee, Ohio University, 2003-2004
Faculty Advisor, Communication Week, Ohio University, 1994-2002
Faculty Advisor, Students in Communication Systems Management, Ohio University, 1994-1996
Member, University General Education Review Committee, Ohio University, 1998-1999
Member, College of Communication Curriculum Committee, Ohio University, 1998-2000
Member, College of Communication Graduate Committee, Ohio University, 1997-2002
Member, University Calendar Review Task Force, Ohio University, 1996-1997
Member, Outstanding Civil Service Award Committee, Ohio University, 1995-1996
Member, Mathematics Department Search Committee, Kenyon College, 1990-1991
Member, Williams Memorial Award Committee, Kenyon College, 1989-1991

Professional Membership

American Economic Association

Ph.D. Dissertation Supervision

“The Examination of Strategic Interactions in One Local Access Telephone Market, the Effects on Expected Price
for Access and Universal Access.” Judith Ann Molka-Danielsen. School of Information Sciences,
Telecommunications Program, University of Pittsburgh, 1998.

Expert Testimony Presented

California (On behalf of The Utility Reform Network [TURN])

PUCC Cause No.                 Title                                  Topic

Applications:                 Review of                               TELRIC Compliance
01-02-024                     UNE Rates                               of UNE Rates.
01-02-035                                                             Progress of local exchange
02-02-031                                                             competition.
02-02-032
02-02-034
02-03-002
(February 7, 2003 [Reply Declaration])
(March 12, 2003 [Rebuttal Declaration])




                                                      6
California (On behalf of The Utility Reform Network [TURN], continued)

PUCC Cause No.               Title                                Topic

Rulemaking                   Permanent                            Pricing and Cost Allocation for the
93-04-003                    Line Sharing                         High Frequency Portion of the
Investigation                Phase II                             Local Loop in the NGDLC
93-04-002                                                         Environment.
(Phase II)
(July, 2001)

Rulemaking                   Permanent                            Pricing and Cost Allocation for the
93-04-003                    Line Sharing                         High Frequency Portion of the
Investigation                Phase I                              Local Loop.
93-04-002
(Phase I)
(June, 2001)

Canadian Radio-Television and Telecommunications Commission
(On Behalf of Action Réseau Consommateur, the Consumers' Association of Canada, Fédération des
Associations Coopératifs d'économie familiale, and the National Anti-Poverty Organization, BC Old Age
Pensioners' Organization, BC Coalition For Information Access, Consumers' Association of Canada (BC
Branch), Council of Senior Citizens' Organizations of BC, End LegislatedPoverty, federatedanti-poverty
groups of BC, Senior Citizens' Association of BC, Tenants Rights Action Coalition, West End Seniors
Network, Consumers’ Association of Canada (Manitoba), Manitoba Society of Seniors, and Manitoba
Keewatinowi Okimakanak Inc.)

CRTC Case No.        Title                                Topic
Public Notice        Price Cap Review                     Price cap regulation and productivity growth.
CRTC 2001-37         and Related Issues                   Accommodative entry policy.
(August, 2001)

Indiana (On behalf of the American Association of Retired Persons and Citizens Action Coalition of
Indiana)

IURC Cause No.       Title                                Topic

42405                SBC Indiana’s Request for            Analysis of local competition,
                     Alternative Regulation               Price Cap Regulation and Productivity
(October, 2003)




                                                  7
Indiana (On behalf of the American Association of Retired Persons and Citizens Action Coalition of
Indiana, continued)

IURC Cause No.      Title                              Topic

41911               Commission’s Investigation         Service Quality Performance.
(July, 2001)        of Ameritech Indiana Service
                    Quality

40785-S1, 40849,    Approval of Settlement             Service quality, Advanced Services

41058               Agreement between Ameritech        Deployment, Alternative Regulation.
(January, 2001)     And other Parties

40785-S1, 40849,    Approval of Settlement             Service quality, Alternative Regulation,
41058               Agreement between Ameritech        Cost of Service, Cost Modeling, Compliance
(August, 2000)      And other Parties                  with §254(k)of the Telecommunications Act of
                                                       1996.

40785-S1            Commission’s Investigation         Economic Cost of Service/
(September, 1999)   Ameritech Indiana’s Compliance     Cost Allocation.
                    With Section 254(k) of the
                    Telecommunication Act

40849
(November, 1997)    Commission’s Own Motion            Interim and Permanent Alternative
                    On Ameritech Indiana’s             Regulation/Rate Design.
                    Request for Interim Relief

40849               Ameritech Indiana                  Interim Alternative Regulation/
(September, 1997)   Request for Interim                Rate Design.
                    Relief

Ohio (On behalf of the Ohio Consumer's Counsel)

PUCO Case Nos.              Title                      Topic

98-1082-TP-AMT              SBC/Ameritech              Sharing of cost saving.
(December, 1998)            Request for Approval       Total factor productivity growth.
                            of Merger



                                                   8
Ohio (On behalf of the Ohio Consumer's Counsel, continued)

PUCO Case Nos.             Title                        Topic

96-899-TP-ALT              Cincinnati Bell              Price Cap Regulation/
(December, 1997)           Alternative                  Rate Rebalancing/
                           Regulation                   Rate Design.

94-2019-TP-ACE             MFS INTELENET                Financial, Managerial,
(May, 1995)                                             and Technical Ability to
                                                        Provide Local Exchange Service.

93-487-TP-ALT and          Ohio Bell:                   Incremental Costs/
93-576-CSS                 Alternative                  Fully Distributed Costs/
(September, 1994)          Regulation                   Alternative Regulation.

Maryland (On behalf of the Maryland People's Counsel)

MPSC Docket No.            Title                        Topic

8730                       Bell Atlantic                ISDN pricing and cost of service.
(Rebuttal Testimony)       ISDN Tariff
(November, 1996)           Proposal

8730                       Bell Atlantic                ISDN pricing and cost of service.
(Direct Testimony)         ISDN Tariff
(October, 1996)            Proposal

8715                       MCI Request                  Price Cap
(Rebuttal Testimony)       for Alternative              Regulation,
(April, 1996)              Regulation for               Cost Allocation and
                           Bell Atlantic                Loop Cost Recovery.
                           Maryland

8715                       MCI Request                  Price Cap
(Direct Testimony)         for Alternative              Regulation,
(March, 1996)              Regulation for               Cost Allocation and
                           Bell Atlantic                Loop Cost Recovery.
                           Maryland




                                               9
Indiana (On behalf of the Indiana Consumer Counselor)
*Testimony prepared, but not filed due to case settlement.

IURC Cause No.              Title                             Topic

40611                       Ameritech Indiana Approval        Analysis of TELRIC studies.
(June, 1997)                of Statement of Generally
                            Available Terms

39853                       Teleport Communications Group     Authority to provide intraLATA
(March, 1994)               of Indiana, Inc.                  and interLATA Private Line
                                                              Services.

39705                       Indiana Bell Telephone            Alternative Regulation/
(January, 1994)                                               Competition/Infrastructure
                                                              Deployment/Imputation.

39474                       Indiana Payphone Association v.   Imputation/separate subsidiary.
(May, 1994)                 Indiana Bell Telephone

39755                       GTE North Inc./GTE                Divestiture of Assets/Policy.
(September, 1993)           Intelligent Network
                            Service Inc.

39718                       Ameritech Advanced Data           Affiliate Relationships.
(August, 1993)              Services

39475                       Indiana Payphone Association      Dial-Around Compensation.
(March, 1993)

38269-S4                    IntraLATA Toll Compensation       Toll Rate Deaveraging.
(February, 1993)

39369                       IURC Investigation into           Access Charge Parity/Recovery
(February, 1993)            Access Charge Parity              of Non-Traffic-Sensitive
                                                              Costs/Policy.

39618                       IURC Investigation into           Collocation Policy.
(January, 1993)             Special Access Collocation

39385                       Indiana Bell Telephone:           Evaluation of Competition in
(October, 1992)             Competition and Pricing           Dedicated Communications
                            Flexibility                       Market/Policy.


                                                  10
Indiana (On behalf of the Indiana Consumer Counselor, continued)

IURC Cause No.                 Title                                  Topic

39353*                         Indiana Gas Company                    Temperature Normalization
                                                                      Tracker/Demand Side
                                                                      Management/Reproduction Cost
                                                                      of Rate Base/Capital Costs.

39314                          Indiana Michigan Power Co.             Clean Air Act Amendments
(September, 1992)                                                     /Demand Side Management.

39221                          American Telecommunications            Financial Viability.
(January, 1992)                Corporation

39215                          Indiana American Water Co.             Reproduction Cost of Rate
(January, 1992)                                                       Base/Capital Costs.

39166                          Indiana Cities Water Co.               Reproduction Cost of Rate
(November, 1991)                                                      Base/Capital Costs.

39164/39165                    Ohio Valley Gas Corp.                  Reproduction Cost of Rate
(October, 1991)                                                       Base/Capital Costs.

39017*                         IURC Investigation into                Reproduction Cost of Rate
                               Indiana Bell Earning                   Base/Capital Costs.




Comments Filed

Federal Communications Commission (On behalf of AARP)

In the Matter of Inquiry into High-Speed Access to the Internet Over Cable and Other Facilities. GN Docket No.
00-185, FCC No. 00-355. “Tangled Web: The Internet and Broadband Open Access Policy.” (January 10,
2001).

Indiana Utility Regulatory Commission (On behalf of the Indiana Consumer Counselor)

A Comprehensive Approach to Local Exchange Competition in Indiana (October, 1995).




                                                     11
Comments Filed (Continued)

Indiana Utility Regulatory Commission (On behalf of the Indiana Consumer Counselor, continued)
Comments of the Office of the Office of Utility Consumer Counselor to the Telecommunications Regulatory
Flexibility Committee (1993).

New York Public Service Commission (On behalf of Independent Telephone Companies [NYNEX and
Rochester excluded])

Proceeding on Motion of the Commission to Examine Issues Related to the Continued Provision of Universal
Service and to Develop a Regulatory Framework for the Transition to Competition in the Local Exchange Market:
"Comments on Compensation Arrangements Related to Module 2" (April, 1995).

Maine Public Service Commission (On behalf of Independent Telephone Companies [NYNEX excluded])

Inquiry Into the Provision of Competitive Telecommunications Services (Chapter 280), Docket 94-114: "Reply
Comments to the `Preliminary Proposal for a Revision and Restructuring of the Access Charge Provision of
Chapter 280'" (June, 1995).

Federal Communications Commission (On behalf of the Indiana Consumer Counselor)

Comments of the Indiana Office of Utility Consumer Counselor on the Ameritech Customers First Plan (1993).

Reply Comments of the Indiana Office of Utility Consumer Counselor on the Ameritech Customers First Plan
(1993).
May 2004




                                                     12
                  Attachment TRR-2
                Cost Modeling Issues
Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                              Attachment TRR-2




 1   Discussion of Cost Modeling Issues:

 2   Extracting Loop Costs from the FCC SM.


 3   In this attachment I discuss how I generated cost estimates using the FCC SM for purposes of

 4   comparison with the LoopCAT model. I used the FCC SM, which I downloaded from the FCC’s

 5   web site. The file needed to update the model’s database to reflect current line counts and to

 6   address the DS-0 equivalent issue (see discussion in a later paragraph) is contained in the

 7   electronic workpapers as “linecount.xls”. The user should update the file labeled “Hcpm.mdb”

 8   in the “db” folder using this file. The user should also copy the file

 9   “hcpm_current_inputs_TRR.xls” into the directory where the FCC SM is set up and select this

10   file when choosing the input file during a model run.



11   As I mentioned above, I adjusted the line counts to reflect SBC Ohio’s 2003 levels. The FCC

12   SM includes a line true-up routine as part of its “Cluster Module.” When the true-up option is

13   selected then during the clustering process, the model compares lines reported in the customer

14   location data set with the lines reported in the data base reporting current line counts for each

15   wire center. A true-up factor representing this is then computed and applied.



16   The user-adjustable inputs in the FCC SM with respect to copper cable identify differing cable

17   input costs, stated in material cost per foot, based on cable size, cable gauge, and whether the

18   deployment is underground, buried, or aerial. Values were taken from the corresponding cable

19   gauges from the LoopCAT model. Linear interpolation was used in the few cases where the FCC

20   SM and LoopCAT did not have identical cable sizes. These values are shown in the file:

                                                             1
                                                                     Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                On Behalf of OCC
                                                                                       Case No. 02-1280-TP-UNC
                                                                                              Attachment TRR-2




 1   “hcpm_current_inputs_TRR.xls”



 2   Given that the FCC SM was created for developing the costs of universal service, I extracted the

 3   direct loop costs estimates from the model. In the electronic workpapers are three FCC SM

 4   output files. One, labeled “FCC_Default Scenario.xls”, is output from the FCC web site which

 5   contains the results of an FCC SM run completed by the FCC. In addition, a filed labeled

 6   “FCC_Line_Trueup.xls” contains the results of the model run that I performed that left the FCC

 7   default input assumptions, but eliminated special access and switched access served by circuits

 8   above DS-0 levels and which matched line count values contained in LoopCAT. Finally the file

 9   “Unified_Scenario.xls” includes the line count true-up, the elimination of circuits above DS-0,

10   and applies the proprietary cable costs, the unifying expense factors, discussed below and in the

11   testimony, and FCC’s default fill factors. The process for extracting the direct loop costs was the

12   same with each of these files. The worksheet titled “USF” shows the total cost per line of

13   providing the universal service offering, including 100% of the costs of the local loop. Results

14   are reported on a CLLI™ code basis. Auditing the USF cost cells in this worksheet reveals that

15   the cost total utilizes information contained in another worksheet titled “Investment Input.” As I

16   was only interested in the cost of loops, auditing cells in the “Investment Input” worksheet

17   revealed that the loop costs utilized to generate the overall universal service cost per line

18   included costs for distribution plant, network interface device, concentrator unit, and feeder

19   plant. Further auditing showed that the value for the loop shown in column “HU” of the

20   “Investment Input” worksheet included an adjustment for retail uncollectables. I changed the

21   uncollectable amount to reflect wholesale rather than retail uncollectables. The other


                                                           2
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                Attachment TRR-2




 1   modification that I made was to extract only the direct costs. The direct costs associated with the

 2   loop components are contained in columns DA (distribution plant), DD (concentrator), DQ

 3   (feeder), and CX (NID) of the “Investment Input” worksheet. Modifying the formula in column

 4   HU to target these direct costs rather than the original loaded direct costs generated the cost

 5   estimate for each CLLI™ code.

 6   LoopCAT Adjustments

 7   In this section I will describe the steps associated with modifying the LoopCAT model. As I

 8   discussed in the testimony, I modified the user-adjustable inputs in the LoopCAT model

 9   associated with fill factors, depreciation lives, and capital costs, utilizing the unifying inputs,

10   discussed below, instead of those specified by SBC Ohio. With regard to depreciation lives and

11   capital costs I modified the file “OH 2002 ACF 2004-01-14.xls”, which had been provided by

12   SBC Ohio with its workpapers. The “Inputs” worksheet of this file allows the user to alter

13   assumptions associated with depreciation lives and salvage values. I applied the unifying

14   assumptions associated with depreciation lives, salvage values, and capital costs in this

15   worksheet and then executed the model according to SBC Ohio’s instructions. The resulting file

16   is provided in the workpapers as “Unified_OH_2002_ACF_.xls”. The resulting capital cost

17   factors (which address depreciation, capital costs, and income taxes), were then copied to the

18   LoopCAT model for 2-wire loops (“OH 2w Analog LoopCAT 04-07 FEB04.xls”).

19

20   To modify the LoopCAT’s fill levels I applied values which I believe are consistent with the

21   achieved fills associated with the FCC SM, applying a 50% fill for distribution and a 70% fill for

22   copper distribution.


                                                             3
                                                                    Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                               On Behalf of OCC
                                                                                      Case No. 02-1280-TP-UNC
                                                                                             Attachment TRR-2




 1   With regard to fiber feeder fill rates, I utilized the FCC SM’s fiber feeder fill of 100%.



 2   I also applied 100% IDLC in the LoopCAT run.



 3   The LoopCAT model produces a total cost per line calculation that includes costs for the main

 4   distribution frame (MDF). Given that the FCC SM does not include the MDF I also modified the

 5   LoopCAT results by removing the MDF to provide cost output which was comparable to the

 6   FCC SM.



 7   The results of this modified LoopCAT run are contained in electronic workpaper in the file

 8   labeled “OH_2w_Analog_LoopCAT_roycroft.xls”.



 9   Three other sensitivity runs of the LoopCAT model are included with the electronic workpapers:

10   •       “OH_2w_Analog_LoopCAT_roycroft_SA.xls” uses all unifying assumptions, but leave

11           fiber fills at SBC Ohio’s default level.

12   •       “OH_2w_Analog_LoopCAT_roycroft_SB.xls” uses all unifying assumptions, but leaves

13           fiber fill and UDLC at SBC Ohio’s default levels.

14   •       “OH_2w_Analog_LoopCAT_roycroft_SC.xls” uses all unifying assumptions, but leaves

15           UDLC at SBC Ohio’s default level.




                                                          4
                                                                       Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                                  On Behalf of OCC
                                                                                         Case No. 02-1280-TP-UNC
                                                                                                Attachment TRR-2




 1   Other Unifying Assumptions:

 2   Depreciation Rates and Salvage Values used (from the Virginia Arbitration Order):

 3       Account Number Description                                            Life          Salvage Value

 4       2111            Total Land                                               1
 5       2112            Total Motor Vehicles                                     8             15.0%

 6       2114            Tools and Other Work Equipment                           12             0.0%
 7       2121.1          Buildings - Administrative                               47             2.0%

 8       2121.2          Buildings - Network                                      47             2.0%

 9       2121.3          Buildings - Network Support                              47             2.0%
10       2122            Furniture                                                15             0.0%

11       2123.1          Office Equipment - Office Support                        10             0.0%
12       2123.2          Office Equipment - Office Communication                  7              0.0%

13       2124            General Purpose Computers                                6              3.0%

14       2212            Digital Electronic Switching                             12             2.0%
15       2220            Operator Systems                                         8              0.0%

16       2231            Radio Systems                                            9              -5.0%
17       2232.157        Circuit Equipment - DDS                                  11             1.0%

18       2232.257        Circuit Equipment - Digital Loop Electronic              11             1.0%

19       2232.357        Circuit Equipment - Digital Other                        11             1.0%
20       2232.57         Circuit Equipment - Analog                               3              0.0%

21       2351            Public Telephone Terminal                                7              0.0%
22       2362            Other Terminal Equipment                                 5              0.0%

23       2411            Poles                                                    30            -90.0%

24       2421.22         Aerial Cable - Metallic                                  20            -10.0%
25       2421.822        Aerial Cable Non - Non-Metallic                          25            -10.0%

26       2422.5          Underground Cable - Metallic                             25            -10.0%
27       2422.85         Underground Cable - Non-Metallic                         25            -10.0%

28       2423.45         Buried Cable Exchange Metallic                           20             -5.0%

29       2423.845        Buried Cable Non-Metallic                                25            -10.0%
30       2426.442        Intrabuilding Cable - Metallic                           20             -5.0%

31       2426.462        Intrabuilding Cable - Non-Metallic                       25            -10.0%
32       2441            Conduit System                                           50            -10.0%

33       2690            Intangible Assets                                        3              0.0%




                                                                 5
                                                                 Direct Testimony of Trevor R. Roycroft, Ph.D.
                                                                                            On Behalf of OCC
                                                                                   Case No. 02-1280-TP-UNC
                                                                                          Attachment TRR-2




 1   Capital Costs (from testimony of OCC witness Dr. J. Randall Woolridge):

                                                                      Weighted
                                      Capitalization   Cost             Cost
                                         Ratios*      Rates            Rates
 2      Long-Term debt               23.4%            5.80%            1.36%
 3      Common Equity                76.6%           10.00%            7.66%
 4      Total                        100.0%                            9.02%



 5   Wire Center Reconciliation


 6   The FCC SM and the LoopCAT cost models do not contain a uniform set of CLLI™ codes.

 7   Table TRR-2A, below, shows a reconciliation of the wire centers across the two models:

 8                  Table TRR-2A: Reconciliation
 9                         of Wire Centers
10                 Wire Center in Wire Center
11                 LoopCAT, Not in FCC SM
12                    FCC SM       Not LoopCAT
13                glfroh65         BRKLOHAA
14                mycyoh76         ECLVOHEA

15   While the CLLI™ codes are not uniform across the models, I don’t believe that this deficiency

16   undermines my conclusions which have been based on the 250 wire center that are common

17   across the models. The combined number of lines associated with the wire centers included in the

18   FCC SM, but not in LoopCAT is about 6,700. The count associated with the two CLLI™ codes

19   present in LoopCAT, but not in the FCC SM is less than 1,000 lines.




                                                       6
               Attachment TRR-3
           SBC Wireless Documents
Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                A special
     joint announcement
  from Cingular Wireless,
SBC Communications Inc.
            and BellSouth


                                                                                              September 9, 2003




      Cingular Wireless, SBC Communications and
      BellSouth to Launch FastForwardTM — Latest
      in a New Category of Services That Integrate
      Wireless and Wireline Communications
      I   Innovation Lets Cingular Customers Route Unlimited Incoming Wireless Calls to
          Wireline Numbers, Roll Over Even More Minutes

      I   New Device Works Exclusively With Cingular Wireless Service, Will Be Available
          Starting Oct. 1

      ATLANTA, Sept. 9, 2003 — Cingular Wireless, one of the United States’ premier wireless
      companies with more than 23 million voice and data subscribers, today joined with its parent
      companies, SBC Communications Inc. (NYSE: SBC) and BellSouth (NYSE: BLS), to introduce
      FastForward™ — a groundbreaking new option that marries the convenience of wireless
      service with customers’ wireline phone service.

        The patented FastForward device,         FastForward lets Cingular customers
      which works exclusively with             forward unlimited incoming wireless       “We’re creating truly
      Cingular service, will be available      calls to their landline phone in the      differentiated products that
                                                                                         will fundamentally change how
      beginning Oct. 1 throughout              local calling area without the minutes
                                                                                         our customers access and use
      Cingular’s coverage area, except         counting against their monthly
                                                                                         communications services.”
      for Puerto Rico where it will be         wireless calling plan. The device is
                                                                                           Ray Wilkins, President-
      available later this year. It is being   designed as a cradle to hold a              Marketing and Sales, SBC
      demonstrated for the first time          wireless phone and plugs into an
      today at Morgan Stanley’s 8th Annual     electrical outlet. When the Cingular      “With virtually all of our
      Global Media & Communications            Wireless phone is cradled, calls to the   wireline customers capable of
      Conference being held in Boston.         wireless phone are forwarded to a         being served by Cingular, this
      SBC and BellSouth executives             designated landline phone, while          announcement is significant.”
                                                                                           Dick Anderson, President-
      are scheduled to present at the          the wireless phone’s battery is
                                                                                           Consumer Markets, BellSouth
      conference.                              automatically recharged.
                                                                                                           SBC Communications Inc. | Investor Briefing Page 2




STRATEGIC BENEFITS                                  Shared Sales Channels. In conjunction             HOW FASTFORWARD WORKS:
For Cingular Wireless and its parent                with expanded bundling, the companies have        SIMPLE TO SET UP, EASY TO USE
companies, FastForward is part of a larger          also developed processes and systems to let       Key to the success of all wireless-wireline
initiative to create a new category of products     customers order Cingular Wireless service         integration opportunities is simplicity and ease
that integrate wireless and wireline marketing      through the SBC and BellSouth wireline sales      of use — both key attributes for FastForward.
and services, simplify the calling process and      channels. Cingular sales through wireline         Simply add the landline number to the wireless
change how people communicate.                      sales channels have increased dramatically        phone’s address book and insert the wireless
• FastForward benefits customers by giving          during recent quarters — with second-             phone into the FastForward cradle. An
   them increased convenience and by helping        quarter gross sales four times their levels two   indicator light illuminates on the cradle,
   them conserve wireless minutes — letting         quarters earlier, representing approximately      showing the forwarding service is active.
   them roll over more minutes each month.          15 percent of Cingular’s gross adds.              Now, calls that would have gone to the wireless
   Cingular is the only U.S. wireless carrier to    Shared Voice Mail. Cingular and its parent        phone will ring the designated landline
   offer Rollover, the wireless plan that lets      companies also offer customers the                telephone. To deactivate, just press the
   customers keep their unused monthly min-         convenience of a single voice mail box for        CANCEL button on the front of the device and
   utes. These customer benefits help Cingular      both wireless and wireline calls, giving          remove the wireless phone from the cradle.
   drive customer retention and growth.             customers greater convenience and efficiency.        SBC residential local phone company
• For Cingular and its parents, FastForward                                                           customers who receive a single bill for
                                                    MinuteShareSM. In June, Cingular Wireless,        Cingular Wireless and landline services will
   creates additional marketplace differen-
                                                    SBC Communications Inc. and BellSouth             not have to pay a monthly charge to receive
   tiation by delivering an option not matched
                                                    announced MinuteShareSM, a new service            unlimited call forwarding. Similarly, BellSouth
   by other wireless providers or by other
                                                    enabling SBC or BellSouth residential             will offer unlimited call forwarding when
   wireline telephone providers.
                                                    customers to share a single bucket of             customers sign up for a combined bill and
• FastForward also helps Cingular improve
                                                    wireline long distance (in states where SBC       the choice of two other features. Customers
   capital efficiency, moving minutes from its
                                                    Long Distance is authorized to sell service)      who use other companies for their local
   wireless network to wireline networks.
                                                    and wireless local and long distance minutes.     landline service pay just $2.99 a month to
THE LATEST IN WIRELESS-WIRELINE                     A market trial of the new service has             receive unlimited call forwarding.
INTEGRATION                                         demonstrated how customers use the added             The new FastForward device can be
FastForward is the latest in a series of initia-    convenience of combined minutes on their          purchased as a stand-alone item at all
tives undertaken over the past several months       home and wireless phones. MinuteShare is          Cingular Wireless retail locations for $39.99.
by Cingular Wireless and its parents to deliver     available in select SBC markets today and is      SBC and BellSouth residential customers
on customer expectations that wireless and          expected to launch more widely this autumn.       purchasing new Cingular Wireless service can
wireline communications be integrated.              BellSouth expects to launch the service in        call their consumer service centers and also
   “Through our initiative with Cingular,           the first quarter of 2004.                        purchase the device for the same amount plus
we’re creating truly differentiated products                                                          tax and shipping and handling. Compatible
                                                    FUTURE OPPORTUNITIES
that will fundamentally change how our                                                                wireless phones must be purchased separately.
customers access and use communications             FastForward and these other strategic
                                                    initiatives are driven by Cingular’s extensive       Initially, the FastForward device will be
services,” said Ray Wilkins, president-SBC                                                            compatible with select Nokia and Motorola
marketing and sales. “Integrating wireless          wireline overlap — with about 80 percent
                                                    of its customer base in either the SBC or         wireless phone models. Additional
with core telephone services will differentiate                                                       FastForward-compatible wireless phones,
us from competitors and help us gain and            BellSouth wireline service region — and by
                                                    customer expectations for truly integrated        including Sony Ericsson and Siemens, are
retain customers.”                                                                                    expected to be available as early as Nov. 1.
   “The integration of wireless and wireline        telecommunications services.
has always been important to BellSouth and             Cingular Wireless and its parents continue
its customers. With virtually all of our wireline   to focus resources on opportunities in this
customers capable of being served by                area. For example, there are opportunities
Cingular, this announcement is significant,”        to combine Wi-Fi services with 3G wireless
said Dick Anderson, president-consumer              data network capabilities. And unified
                                                                                                        Cingular Focus
markets, BellSouth.                                 messaging — combined wireline voice mail,
   In addition to FastForward, major steps the      wireless voice mail and e-mail in a single          • 236 million licensed POPs
companies have taken include:                       message box — holds the promise of giving           • $3.8 billion revenue
                                                    customers communications management                 • $756 million operating income
Wireline-Wireless Bundles. Both SBC                 tools that seamlessly connect voice and
Communications Inc. and BellSouth have                                                                  • Ownership: 60 percent SBC,
                                                    data platforms on both wireline and wireless                        40 percent BellSouth
introduced mass market service bundles              networks.
that provide conveniences such as one-stop
shopping, a single bill and discounts for both
Cingular wireless and wireline services.
                                                                                                                   SBC Communications Inc. | Investor Briefing Page 3




ABOUT CINGULAR WIRELESS
Cingular Wireless, a joint venture between SBC Communications (NYSE - SBC) and BellSouth (NYSE - BLS),
serves more than 23 million voice and data customers across the United States. A leader in mobile voice
and data communications, Cingular is the only U.S. wireless carrier to offer Rollover, the wireless plan that
lets customers keep their unused monthly minutes. Cingular has launched the world’s first commercial
deployment of wireless services using Enhanced Data for Global Evolution (EDGE) technology. Cingular
provides cellular/PCS service in 43 of the top 50 markets nationwide, and provides corporate e-mail and
other advanced data services through its GPRS, EDGE and Mobitex packet data networks. Details of the
company are available at www.cingular.com.

ABOUT SBC COMMUNICATIONS
SBC Communications Inc. (www.sbc.com), through its subsidiaries, is one of the world’s leading data,
voice and Internet services providers. Through their world-class networks, SBC companies provide a full
range of voice, data, networking and e-business services, as well as directory advertising and publishing.
A Fortune 30 company, SBC Communications Inc., through its subsidiaries, is America’s leading provider
of high-speed DSL Internet Access services and one of the nation’s leading Internet Service Providers.
SBC companies currently serve 56 million access lines nationwide. In addition, SBC companies own
60 percent of America’s second-largest wireless company, Cingular Wireless, which serves more than
23 million voice and data wireless customers over its nationwide voice and data networks. Cingular was
the first company in the United States to commercially launch EDGE technology. Internationally, SBC             Contact Information
companies have telecommunications investments in 22 countries.
                                                                                                                Cingular Wireless . . . . . . 404-236-6203
ABOUT BELLSOUTH CORPORATION                                                                                     Kent Evans,
BellSouth Corporation is a Fortune 100 communications services company headquartered in Atlanta, Ga.              Executive Director-Investor Relations
BellSouth serves more than 45 million local, long distance, Internet and wireless customers in the
United States and 14 other countries.
                                                                                                                BellSouth
   Consistently recognized for customer satisfaction, BellSouth provides a full array of broadband data
solutions to large, medium and small businesses. In the residential market, BellSouth offers DSL high-speed
                                                                                                                  Investor Relations . . . 404-249-3491
Internet access, advanced voice features and other services. BellSouth also offers long distance service        Nancy Davis,
throughout its markets, serving both business and residential customers. The company’s BellSouth                  Vice President
AnswersSM package combines local and long distance service with an array of calling features; wireless
data, voice and e-mail services; and high-speed DSL or dial-up Internet service. BellSouth also provides        SBC Investor Relations . . 210-351-3327
online and directory advertising services through BellSouth® RealPages.comSM and The Real Yellow Pages®.        Mike Coffey,
BellSouth owns 40 percent of Cingular Wireless, the nation’s second-largest wireless company, which               Managing Director
provides innovative data and voice services.
                                          Wireless
                                             Momentum
SBC Communications Inc. and Cingular Wireless – At a Glance
Wireless is a key component in the growing SBC portfolio of                    SBC Communications Inc., which owns 60 percent of Cingular
bundled communications offerings for both consumers and                        Wireless through a joint venture with BellSouth, has been
businesses. Wireless service enhances the overall value of                     offering bundled wireline-wireless packages of services
SBC bundles by allowing customers to choose from a range of                    through its local service companies since early summer 2002.
national and local wireless calling plans, offering them at a                  SBC companies work with Cingular, the nation’s second
significant discount and delivering on customers’ desire for                   largest wireless company, to bring innovative calling bundles
convenient and simplified billing, ordering and care.                          and integrated wireline-wireless services to market.


                                                                                                 SBC Communications Inc. and Cingular
       SBC United States Inc.: 90 Percent of SBC
     SBC CommunicationsOperations
                                                                                                 have one of the largest wireline-wireless
     Customers Live in the Cingular Footprint
                                                                                                 service- area overlaps.
                                                                                                 • Cingular is the nation’s second-largest
                                       236 million POPs,
                                       industry’s largest                                          wireless provider.
                                       wireless-wireline                                         • SBC companies serve one-third of the
                                           overlap
                                                                                                   nation’s wireline customers.

                                                                                                 • Cingular provides service in SBC states,
                                                                                                   which include California, Arkansas,
                                                                                                   Kansas, Missouri, Oklahoma, Texas,
                                                                                                   Connecticut, Illinois, Indiana, Michigan,
                                                                                                   Nevada, Ohio and Wisconsin.

         HAWAII
                                                                                                 • Nearly half of Cingular’s 23 million
                                                                                                   customers currently reside in
         O P E RAT I O N S                                                                         SBC territories.
         nn United States Operations




     SBC Communications Inc. – Accelerating Sales of Wireline-Wireless Bundles
                                                   Compelling Wireless-Wireline Bundles,         • SBC consumer service representatives
     SBC Wireline Packaged Sales                   Strong Sales Channels Boost Wireless Sales      receive millions of calls every month,
     of Cingular                                                                                   offering opportunities to cross-sell
     (In thousands)                    183         • SBC companies continue to increase
                                                                                                   wireless.
                                                     sales of new wireless bundles. In the
                                                     second quarter of 2003, SBC sales           • Integrated ordering systems allow a
                                                     channels sold 183,000 wireless bundles,       SBC customer service representative to
                                                     up from 68,000 the quarter before.            handle wireline and wireless orders,
                                                                                                   answer questions and make upgrades.
                                                   • Today, SBC companies are the No. 1
                             68                                                                  • Integrated billing systems allow
                                                     sales agents for Cingular postpaid sales.
                                                                                                   customers to receive one SBC monthly
        32                                         • SBC and Cingular account teams are
                                                                                                   bill for both wireline and wireless
                                                     working collaboratively to sell to
                                                                                                   services.
                                                     enterprise customers.
      4Q/02 1Q/03 2Q/03


                                                                                                                                            9/5/03
                           Wireless
                           Innovative Offerings, Value,
                               Choice and Convenience

MinuteShareTM – Erasing the Distinction Between Wireless and Wireline Calls
                                             SBC Communications Inc. and Cingular Wireless have rolled out a new service in select
                                             markets called MinuteShareTM, enabling residential consumers to share a single bucket
                                             of minutes for calls made from either their SBC local wireline or their Cingular
                                             wireless phones.

                                             MinuteShare, the only service of its kind available in the marketplace, includes a
                                             single, shared bucket of wireline long distance and wireless local and long distance
                                             minutes, giving customers more flexibility in using their minutes. The service is
                                             designed for consumers who use the large number of night and weekend minutes
                                             typically found in wireless plans to make long distance calls from home.

                                             With MinuteShare, they can take advantage of these minutes to make long distance
                                             calls while enjoying the clarity and quality of their home wireline phone. Subscribers
                                             who already have unlimited local service continue to have unlimited local calling from
                                             their home wireline phone and will enjoy the convenience of a single bill and single
                                             customer service contact.


Integrating Wi-Fi and 3G Service
SBC Communications Inc. will use its relationship with Cingular to bring
to market a service that combines the strengths and benefits of Wi-Fi hot
spots and 3G wireless data services. The service will allow subscribers to
move between their home or office SBC broadband service, SBC
FreedomLink Wi-Fi hot spots and Cingular’s network while experiencing
not only speed and mobility, but also service consistency and simplicity.

To provide true 3G wireless data services and speeds, Cingular is currently
deploying Enhanced Datarate for Global Evolution (EDGE) technology.
EDGE is a software enhancement to Cingular’s General Packet Radio
Service (GPRS) data network. To deliver an integrated Wi-Fi and 3G service,
SBC companies and Cingular are working on a solution to allow roaming
between home and office LANs, Wi-Fi hot spots, and the Cingular
GPRS/EDGE network so that subscribers can experience broadband
Internet access regardless of where they are.

Users could experience speeds between 2 megabits per second and 5 megabits per second between access points and their Wi-Fi-
enabled devices on the SBC FreedomLink Wi-Fi network and up to 170 kilobits per second on Cingular's EDGE network. The
integrated service is expected to be available in late 2004 or early 2005.



                                                                                                                                     9/5/03
                         Wireless
                         Innovative Offerings, Value,
                             Choice and Convenience
SBC wireline-wireless bundles and packages provide new Cingular customers with a range of discounts on
monthly recurring charges – from 5 percent to 20 percent – depending on the plan and package selected.


            Select monthly national wireless plan rates, including nationwide calling
     5%
      off   packages featuring 250, 350 or 500 anytime minutes and 5,000 night
            and weekend minutes. Available for wireline customers with combined billing.


            Select monthly home calling plans, including packages featuring 300, 400 and
    10%
      off   600 anytime minutes and 5,000 night and weekend minutes. Available for
            wireline customers with combined billing.


            Select monthly home and national calling plans that include at least
    20%
      off   250 anytime minutes and 5,000 weekend minutes, when purchased as part of a
            SBC Total Connections plan for both residential and small-business customers.


            For existing Cingular customers who agree to a combined SBC bill and subscribe
     $3
      off   to SBC Total Connections.




    To Order

    In Texas, Oklahoma, Missouri, Arkansas or Kansas, call 1.800.299.3810 or visit
    www.sbc.com for additional details.

    In Illinois, Michigan, Wisconsin, Ohio or Indiana, call 1.800.870.6361 or visit
    www.sbc.com for additional details.

    In California or Nevada, call 1.800.310.2355 or visit www.sbc.com for additional
    details.

    In Connecticut, call 1.800.419.4381 or visit www.sbc.com for additional details.




                                                                                                         9/5/03
                  Attachment TRR-4
                 Discovery Responses
Direct Testimony of Trevor R. Roycroft, Ph.D.
                           On Behalf of OCC
                  Case No. 02-1280-TP-UNC
                                 Public Utilities Commission of Ohio
                                     Case No. 02-1280-TP-UNC
                                  SBC OHIO’S RESPONSES TO
                          Ohio Consumers’ Counsel 3rd Set of Data Requests


Interrogatory # 102:

Referring to IURC 42393 Order, page 43, and the discussion of loops that terminate without
distribution cable:

a)       Has the LoopCAT model filed by SBC Ohio been modified to incorporate loops that are
         terminated without distribution?

b)       If the LoopCAT model has been modified, where in the model has this modification be
         undertaken? Please identify specific worksheets and cells, and any other related files.



Response:

(a-b)    No.



Responsible Person:             James R. Smallwood
                                Associate Director-Cost Analysis
                                One SBC Center, Room 38-X-8
                                St. Louis, MO 63101




CHDB04 13169012.1 29-Apr-04 15:28                   17

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:22
posted:3/1/2012
language:
pages:163