|Federal Communications Commission
1919 - M Street, N.W.
Washington, D.C. 20554
|News media information 202 / 418-0500
Fax-On-Demand 202 / 418-2830
Released: November 13, 1997
GUIDANCE TO PROPONENTS OF COST MODELS
In the Universal Service Order released May 8, 1997, the Commission, acting on the recommendation of the Federal-State Joint Board, concluded that non-rural carriers providing supported services to rural, insular, and high cost areas (collectively referred to as high cost areas) should receive universal service support based on the forward-looking cost of providing the supported services.(1) The Commission determined that it could not select a mechanism for computing forward-looking costs because none of the mechanisms that had been submitted for consideration was sufficiently developed at that time.(2) The Commission stated that it would continue to review two cost models, the Hatfield Model and the Benchmark Cost Proxy Model (the BCPM).(3) The Commission stated that it would select a forward-looking economic cost mechanism with platform design features and input values by August 1998.(4)
In a Further Notice of Proposed Rulemaking in this proceeding (FNPRM), the Commission established a multi-step approach to refining and selecting a mechanism for determining a non-rural carrier's forward-looking economic cost of providing supported services to high cost areas.(5) The Commission specified that the Common Carrier Bureau (Bureau) would "issue orders and public notices on a regular basis explaining its analysis of the model submissions and industry comments and to select particular design features."(6) The Commission further stated its expectation that "such guidance from the Bureau will provide the proponents with necessary direction to refine their models."(7)
In meetings with proponents of the BCPM and the Hatfield model and other interested parties, the Bureau staff has encouraged the continued evolution and refinement of the models.(8) This Public Notice offers guidance to proponents of models regarding the customer location and outside plant platform design issues raised in the FNPRM.(9) In the Order, the Commission stated its "anticipat[ion] that by the end of the year [it] will choose a specific model" as the platform for a federal mechanism. In order to choose a "specific model," however, the Commission must evaluate and compare completed versions of the models.(10) The Bureau therefore requests that parties seeking consideration of their model as the platform for a federal mechanism submit their models within four weeks after the release of this Public Notice. Models that conform to the guidance in this Public Notice are likely to be considered more favorably in this proceeding.
The Commission stated in the FNPRM that it may select a model submitted to the Commission by a proponent, or it may select a hybrid model incorporating the best features of proposed models and design components developed by the Commission staff or other parties.(11) On October 31, 1997, staff members of the Common Carrier Bureau proposed an alternate approach to customer location and outside plant design issues in the form of a Hybrid Cost Proxy Model (HCPM).(12) The HCPM is in many respects a hybrid of the BCPM and the Hatfield model, although it also contains features that differ from both the BCPM and the Hatfield model. The Bureau anticipates that the Commission will consider the HCPM as an alternative to the customer location and outside plant design modules in the BCPM and the Hatfield model. The Bureau observes, however, that the FNPRM leaves open the possibility that the Commission may consider other models or other components of models in selecting the best mechanism for determining non-rural carriers' forward looking cost for providing the supported services.
I. Customer Location
A. Geocode Data
The FNPRM requested comment on the use of data that associate the location of each customer with latitudinal and longitudinal coordinates (geocode data) in a forward-looking economic cost mechanism.(13) Many commenters agree that geocode data, which provide the actual geographic location of customers, are preferable to algorithms intended to estimate customer locations based on Census data or other information regarding the number of customers in a given geographic area.(14) Because assumptions about the location of customers have a large impact on loop length calculations, the use of more accurate customer location data is consistent with the criterion specified in the Order that "a model's average loop length should reflect the incumbent carrier's actual average loop length."(15) Some commenters, however, question the feasibility of using geocode data in the federal mechanism because of the lack of reliable data in rural areas and the burden of developing such data.(16) The Bureau recommends that models be capable of accepting and using geocode data to the extent that such data are available and reliable.
B. Wire Center Boundaries
In their evaluation of previous versions of cost models submitted to the Commission, State members of the Joint Board have noted that inaccurately mapping customers to a wire center may result in inaccurate line counts and impede the determination of the most efficient engineering practices for serving that wire center.(17) Through the model development process, the BCPM and the Hatfield model have been refined so that both models determine wire center boundaries based, at least in part, on a database provided by Business Location Research (BLR).(18) The BCPM has improved the accuracy of the BLR data it uses by using wire center boundary data based on Census Blocks (CBs), rather than the larger Census Block Groups (CBGs).(19) The Hatfield model also uses BLR data at the Census Block-level, although its proponents have stated that they intend to use the Census Block Group data in isolated instances where those data appear to be more accurate.(20) This refinement decreases the discrepancies between the two industry-proposed models. Consistent with the criterion specified in the Order that "[w]ire center line counts should equal actual ILEC wire center line counts" however, the Bureau recommends that models be capable of accepting wire center boundary data in standard Geographic Information System (GIS) format from any source that the Commission finds may estimate those boundaries more accurately.(21)
II. Outside Plant Design
A. Documentation of Assumptions
The Order requires that cost models employ the "least-cost, most-efficient and reasonable technology for providing the supported services that is currently being deployed," and that all engineering assumptions be reasonable.(22) Furthermore, the Order also requires the models' algorithms, data, and assumptions to be open and verifiable.(23) These criteria suggest that outside plant design should be considered both from an engineering perspective, to ensure that the network provides the type and quality of service specified in the Order, and from an economic perspective, to ensure that the network design minimizes cost and maximizes efficiency. Moreover, the requirement that assumptions be open and verifiable ensures that the Commission can confirm that the other criteria have been met. To the extent that models' algorithms do not explicitly explore different loop architectures in varying situations and select the least-cost alternative for that particular situation, the Bureau recommends that model proponents provide detailed documentation that explains and justifies any assumptions and engineering rules of thumb that their models employ. This documentation should demonstrate how these assumptions and rules of thumb meet the Order's requirement that a model employ the least-cost, most-efficient, and reasonable technology.
One assumption for which model proponents should provide documentation is a model's algorithm for deploying digital loop carrier (DLC) devices. For example, the basic outside plant design structures presently employed in the BCPM and the Hatfield model involve running optical fiber feeder cables from the central office to a point within designated serving areas and serving the customers within each serving area with copper distribution cables from a DLC device within the serving area. Because DLC devices are expensive, costs per customer can be minimized by connecting larger numbers of customers to each device, subject to the DLC's capacity limitations and the limits on the length of the copper cable between the DLC and the customer premises. The BCPM and the Hatfield model currently determine which customers are located in a given serving area using either grids (BCPM)(24) or clusters (Hatfield). Both approaches must account for the fact that, particularly in rural areas, some customers are located relatively far from other customers, and therefore are difficult to associate with any single serving area. Presently, the Hatfield model serves geographically isolated customers with the nearest DLC, while the BCPM will place a separate DLC in grids with only a small number of customers. The Bureau recommends that proponents of a model demonstrate how their approaches to deploying DLC devices employ the least-cost, most-efficient, and reasonable technology, as required by the Commission's Order.(25)
Another algorithm that is relevant to whether a model has employed the least-cost, most-efficient, and reasonable technology is a model's algorithm for feeder routes. Earlier versions of the BCPM and the Hatfield model extended four fiber feeder cables, at 90 degree angles from each other, from each central office in all cases. More recent iterations of these models have eliminated feeder cables to quadrants with no population and adjusted feeder route angles directing feeder cables to areas of population concentration. Although these approaches seem to represent improvements in the designs of these industry-proposed models, the Bureau recommends that each proponent of a model demonstrate how their feeder routing algorithms meet the criterion of being the least-cost, most-efficient, and reasonable technology currently being deployed.
In addition to the examples of DLC placement and feeder routing, the Bureau recommends that model proponents demonstrate how every aspect of their outside plant design approach is consistent with the least-cost criterion, while maintaining the network standards established in the Order.
B. Advanced Services
The Commission specified in the Order that the loop design in a forward-looking mechanism "should not impede the provision of advanced services."(26) For example, the Commission determined that loading coils may impede advanced services such as high-speed data transmission and therefore disallowed their use in a model.(27) The Bureau recommends that model proponents explain their assumptions about network configurations and capacity, and explain why such assumptions are reasonable and consistent with common configurations and capabilities of networks of non-rural carriers. For example, model proponents should demonstrate how their models permit standard customer premises equipment (CPE) available to consumers today, such as 28.8 kbps or 56 kbps modems, to perform at speeds at least as fast as the same CPE can perform on the typical existing network of a non-rural carrier.(28)
The Commission also concluded that the definition of supported services should "advance with technology" and will be re-examined in light of "changes in technology, network capacity, consumer demand, and service deployment."(29) The Bureau therefore recommends that models incorporate sufficient flexibility in their loop design algorithms so that the platform of the selected model does not have to be rebuilt in the event that the Commission revises the definition of universal service.
C. Wireless Threshold
In the FNPRM, the Commission sought comment on whether a model should assume that, if the loop investment for a single customer exceeds a certain threshold, an efficient carrier would substitute wireless service for wireline service.(30) The Commission's directive that a cost model use the "least-cost, most-efficient, and reasonable technology" suggests that a model should be able to use information about the costs of wireless service if the Commission concludes that such data are available and reliable.(31) Because the Commission also determined that support calculations should be based on a geographic area that is the size of a wire center or smaller, and the geographic area for estimating costs may not be larger than the support area, the Bureau recommends that models be capable of accomodating as inputs wireless cost thresholds at the level of the wire center or a smaller geographic unit.(32)
D. Fiber-Copper Cross-Over Point
The fiber-copper cross-over point determines where the network will employ optical fiber cable rather than copper cable in its feeder plant. The Commission specified in the Order that a model "must include the capability to examine and modify the critical assumptions and engineering principles . . . includ[ing] . . . fiber-copper cross-over points."(33) While the BCPM assumes that the maximum copper loop length may be 12,000 feet and the Hatfield model assumes that the maximum copper loop length may be 18,000 feet, the Commission noted in the FNPRM that neither proponent has documented that its assumption is the least-cost alternative.(34) In order for the Commission to better understand the cost differences associated with each of these assumptions, the Bureau recommends that proponents of models provide comparative outputs for each of the following five states, using both the 12,000 foot standard and the 18,000 foot standard: Florida, Georgia, Maryland, Missouri, and Montana.
E. Proprietary or Confidential Information
In light of the Commission's requirement in the Order that "all underlying data, formulae, computations, and software associated with the model must be available to interested parties for review and comment," the Bureau recommends that each model proponent submit detailed descriptions of all information or software alleged to be confidential, proprietary, or otherwise unavailable to the public that is used either in the model or in a preprocessing module.(35) The descriptions should include estimates of the costs and procedures that may be associated with making the information or software available to the Commission and to the administrator of the universal service support mechanisms.(36)
III. Follow-Up Requirements
The Commission established criteria for its forward-looking economic cost mechanism in the Order.(37) The Bureau recommends that model proponents ensure that their modules for determining the location of customers and estimating outside plant investment comply with all of the criteria set out in the Order, in addition to the recommendations in this Public Notice.
The Bureau recognizes that proponents of models may need to make certain changes to
their models to bring them into conformity with the guidance provided in this Public Notice.
Within four weeks from the release date of this Public Notice, any proponents of models should
submit their models for consideration by the Commission. To facilitate that process and the
Commission's review, models should be accompanied by
IV. Procedural Matters
Within four weeks of the release date of this Public Notice, proponents of a model should
file an original and three (3) copies of their submission, referencing CC Dockets Nos. 96-45 and
97-160, with the Office of the Secretary, Federal Communications Commission, 1919 M Street,
N.W., Room 222, Washington, DC 20554. Proponents should also provide four (4) copies of
their submission to Chuck Keller of the Universal Service Branch, 2100 M Street, N.W., Room
8918, Washington, D.C. 20554.
Questions regarding this Public Notice should be directed to Chuck Keller (firstname.lastname@example.org) 202/418-7380; Bob Loube (email@example.com) 202/418-7379; or Natalie Wales (firstname.lastname@example.org) 202/418-7389.
Action by the Chief, Common Carrier Bureau.
1. Federal-State Joint Board on Universal Service (Joint Board), CC Docket No. 96-45, Report and Order, FCC 97-157 (rel. May 8, 1997) (Order) at para. 224.
2. Order at para. 245.
3. The proponents of the Hatfield Model are AT&T and MCI. The proponents of the BCPM are US West, Sprint, and BellSouth. See Order at Appendix J for a description of the Hatfield Model and the BCPM.
4. Order at para. 245. In the context of a forward-looking economic cost mechanism, the "platform" refers to the fixed algorithms and assumptions built into a cost model, as contrasted with user-specified "inputs" into a cost model. See Federal-State Joint Board on Universal Service, Forward Looking Mechanism for High Cost Support for Non-Rural LECs, CC Docket Nos. 96-45 and 97-160, Further Notice of Proposed Rulemaking, FCC 97-256 (rel. Jul. 18, 1997) (FNPRM) at paras. 17-18.
5. FNPRM at para. 35.
6. FNPRM at para. 26.
7. FNPRM at para. 26.
8. See Weekly Meetings on Forward-Looking Cost Mechanisms for Universal Service Support, Public Notice, DA 97-1810 (rel. Aug. 22, 1997); October 1, 1997 Meeting on Forward-Looking Cost Mechanism for Universal Service Support for Non-Rural Carriers Rescheduled to September 30, 1997, Public Notice, DA 97-2050 (rel. Sept. 23, 1997).
9. FNPRM at paras. 39-120. In a previous Public Notice, the Bureau provided guidance on the platform issues relating to switching, interoffice trunking, signaling, and local tandem investment raised in the FNPRM. Guidance to Proponents of Cost Models in Universal Service Proceedings: Switching, Interoffice Trunking, Signaling, and Local Tandem Investment, Public Notice, DA 97-1912 (rel. Sept. 3, 1997).
10. Order at para. 245.
11. FNPRM at para. 35.
12. Public Notice, Common Carrier Bureau Makes Available Potential Modules for Determining Customer Location and Outside Plant Design in Forward-Looking Mechanisms for Determining Universal Service Support for Non-Rural Carriers, DA 97-2311 (rel. Oct. 31, 1997). The HCPM is the work of individual staff members, and does not in any way represent the view of the Commission or any given Commissioner.
13. FNPRM at paras. 44, 176.
14. See Aliant Sep. 2 comments at 2; Ameritech Sep. 2 comments at 6; AT&T Sep. 2 comments at 7-8; RUS Sep. 2 comments at 2; AT&T Sep. 10 reply comments at 12-13.
15. See FNPRM at para. 44; Order at para. 250, criterion 1.
16. See GTE Sep. 2 comments at 11-12; Bell Atlantic Sep. 10 reply comments at 3; GTE Sep. 10 reply comments at 4-5; SBC Sep. 10 reply comments at 6-7.
17. State Members' Report on the Use of Cost Proxy Models, Mar. 26, 1997, app. A at 8.
18. BellSouth et. al. Sep. 2 comments at 20; Letter from Richard N. Clarke, AT&T, to William Caton, FCC, dated Oct. 30, 1997 (AT&T Oct. 30 ex parte).
19. Letter from Glenn Brown, U S WEST, to William Caton, FCC, dated Aug. 28, 1997, Draft BCPM Development Notes at 5.
20. AT&T Oct. 30 ex parte at 1.
21. See Order at para. 250, criterion 1. A GIS is a standardized system for storing and manipulating spatially-referenced data.
22. Order at para. 250, criteria 1 and 8.
23. Order at para. 250, criteria 8 and 9.
24. The HCPM also uses grids.
25. Order at para. 250, criterion 1.
26. Order at para. 250, criterion 1.
27. See Order at para. 250, criterion 1; FNPRM at note 137.
28. The Bureau understands that many current transmission networks may limit modem speeds to some extent, depending on the particular network configuration.
29. Order at para. 64.
30. FNPRM at paras. 98-102.
31. See Order at para. 250, criterion 1.
32. See Order at para. 250, criterion 10.
33. Order at para. 250, criterion 9.
34. FNPRM at para. 85.
35. Order at para. 250, criterion 8.
36. Proponents of the BCPM have submitted information that appears to fulfill this recommendation. Letter from Warren D. Hannah, Sprint, to William Caton, FCC, dated Oct. 24, 1997. Proponents of the Hatfield model have submitted a list of data sources and a description of their use in the model. Letter from Richard N. Clarke, AT&T, to William Caton, FCC, dated Oct. 21, 1997. If compliance with the recommendations in this Public Notice changes the accuracy of these submissions, proponents should submit revised estimates.
37. Order at para. 250. To ensure consistency in calculations of federal universal service support, any cost model or cost study, including a state-submitted cost study, must meet the following criteria: