980120.docx

MEMORANDUM
TO:MYRNA J. WALTERS, COMMISSION SECRETARY
INTERESTED PARTIES
FROM:WELDON B. STUTZMAN, DEPUTY ATTORNEY GENERAL
DATE:JANUARY 20, 1998
SUBJECT:SUBMISSION OF INITIAL COST MODEL TESTIMONY
CASE NO. GNR-T-97-22 COST MODEL NOTICE
As agreed at the prehearing conference, Staff offers the following guidelines for all parties preparing to present a particular model for consideration by the Commission.  These guidelines apply also to parties merely wishing to offer testimony rather than formal models.  The initial aim is to narrow the range of issues and inputs to be considered in comparing the various models.  Running with these common inputs will help to isolate the important differences among the models.
This proceeding is not about defining the magnitude of the universal service obligation in the State of Idaho, but about choosing a model to assist in that process.  Ideally, the process of model selection will concentrate on the structure of the models themselves and their ability to aid this Commission in defining the costs of various universal service policy obligations.  Staff believes it worthwhile at this juncture to attempt to evaluate the models on their own merits, rather than on the final size of the USF obligation each model is used to generate.  The final USF obligation in the State of Idaho depends heavily on policy choices about
the number and type of lines to be supported, the benchmark support level, and the size of service area to be considered, perhaps not primarily on the model used to make the final estimation.  Every attempt should be made at this stage to concentrate on the models themselves.
1.  Common inputs.  The initial model runs should be made with the items listed below as common inputs.  Staff’s purpose here is to narrow the range of different assumptions among the models so that we can concentrate attention on those variables and assumptions that represent critical differences among the various models under consideration.  These inputs may or may not be at exactly the right level, a subject that will ultimately have to be decided before any model is utilized to calculate a required level of USF funding for the State of Idaho or for any particular company.  The list of common inputs is as follows:
DEFAULT INPUTS for GNR-T-97-22
(based largely on USW-S-96-5)
Item  No.
Input  Category
Input  Value
1
Cost of debt
7.3%
2
Debt fraction
44.0%
3
Cost of equity
11.2%
4
Rate of Return
9.4%
5
Depreciation:
Economic Life
Net Salvage
Years
Percent
 Motor vehicles
8.2
11.0%
 Garage work equipment
15
0.0%
 Other work equipment
15
0.0%
 Buildings
42
-1.0%
 Furniture
17.5
0.0%
 Office support equipment
10
0.0%
 Company communication equipment
5.5
0.0%
 General purpose computers
6
3.0%
 Digital electronic switching
10
0.0%
 Operator systems
5
0.0%
 Digital circuit equipment
10
4.0%
 Public telephone terminal equipment
7
5.0%
 Poles
18.5
-109.0%
 Aerial cable-metallic
15
-25.0%
 Aerial cable-non metallic
20
-25.0%
 Underground cable-metallic
15
-20.0%
 Underground cable-non metallic
20
-20.0%
 Buried cable-metallic
20
-4.0%
 Buried cable-non metallic
20
-4.0%
 Intrabuilding cable-metallic
20
-1.0%
 Intrabuilding cable-non metallic
20
-1.0%
 Conduit systems
60
-4.0%
6
Tax rates
 Federal income tax
35%
 State income tax
8%
 Property tax
Average State Rate
1.447%
 Gross receipts tax
PUC Assessment
.2348%
2.  Wire centers.  Models should be run for all U S WEST southern Idaho wire centers.  The specific wire centers intended to be used for comparison purposes are Boise-Main [BOISIDMA], Pocatello-North [PCTLIDNO], American Falls [AMFLIDMA], and Castleford [CSFRIDMA].  Since much discussion centers around the difficulty of creating models to accurately depict the costs of serving small and dispersed users, a fourth wire center has been proposed.  This affords the opportunity to compare and contrast modeled cost levels for a very small and likely high cost wire center.  Model proposers should also provide the statewide total costs associated with running the above wire centers.
3.  Filing of results.  By the February 17 date specified in the Order, parties should file model runs using the common inputs and wire centers specified above.  Each model filing should contain a CD with the model run, as well as a hard copy of the final outputs.  The hard copy will assure that Staff and other intervenors have a proof copy of results to guide their own efforts in replicating the modeling done by the proponent.  Formal documentation and summary description of each model should accompany the model runs.
In addition, the model filing should be accompanied by testimony in question and answer format.  The testimony should be limited in size and should concentrate on the major drivers of the level of cost arrived at by the model.  (Staff suggests an informal limit of
twenty-five pages and about ten drivers.)  Testimony should attempt to avoid arcane technical arguments and aim to explain in layman’s language what are the most important variables (either formal input values or assumptions about model structure and operation) that affect the modeled level of loop cost.  Model proponents should explain how and why they made the particular choices that distinguish their model from others.  This evidence is most vital to the practical  ability of this Commission to make an informed choice.
Both for formal filing of model and for the accompanying testimony, the aim is providing judiciously selective information to assist this Commission and Staff in understanding what is vital to the structure and operations of the various models.  Voluminous filings of highly technical material will serve little purpose in this proceeding.  Staff asks that all parties to the case act with civility in asking and providing informal responses to questions about model specifics where they arise but that parties generally refrain from presenting such
all-encompassing details in favor of judgments about limited areas of vital importance.
4.  Additional items.  Staff requests one additional piece of testimony not formally discussed in the prehearing conference.  There is ample room for confusion about how to choose cost models.  Some believe it is improper to choose a model based on the realism of its assumptions.  Others believe it is equally improper to choose whatever model seems to bring about some “properly-sized” universal service obligation.  This is a difficult choice on which the Commission deserves explicit guidance.  Staff requests each party to the case provide its own specific and practical advice on how it believes the Commission should go about choosing a model.  Long lists of generalities and platitudes will not be helpful here.  What is sought is a pointed answer to the question of what determines whether a cost model is appropriate for use in Idaho.
Staff believes there is a case to be made for a second run of all contending models sometime after the March 9 hearing.  Evidence presented at the hearing will likely provide further elucidation of the key variables to be considered.  Based on that evidence Staff may choose to further specify certain variables or to expand the original list of defaults.  A further run of all models subsequent to the hearing would likely be helpful in reaching final conclusions about which model is most appropriate for use in Idaho.  Somewhere around March 25 would be a likely date.
beastla/wpfiles/cost1.wpd