********************************************************
NOTICE
********************************************************
This document was converted from
WordPerfect to ASCII Text format.
Content from the original version of the document such as
headers, footers, footnotes, endnotes, graphics, and page numbers
will not show up in this text version.
All text attributes such as bold, italic, underlining, etc. from the
original document will not show up in this text version.
Features of the original document layout such as
columns, tables, line and letter spacing, pagination, and margins
will not be preserved in the text version.
If you need the complete document, download the
WordPerfect version or Adobe Acrobat version, if available.
*****************************************************************
Before the
Federal Communications Commission
Washington, D.C. 20554
In the Matter of)
)
National Exchange Carrier )
Association, Inc. (NECA) ) AAD 97-2
Proposed Modifications to the 1997 )
Interstate Average Schedule Formulas )
and
)
Proposed Further Modifications to the 1997-98 ) AAD 97-109
Interstate Average Schedule Formulas )
ORDER ON RECONSIDERATION
AND
ORDER
Adopted: December 24, 1997 Released: December 24, 1997
By the Chief, Common Carrier Bureau:
I. INTRODUCTION
1. On December 31, 1996, the National Exchange Carrier Association, Inc.
("NECA") submitted its annual proposed revisions to its average schedule formulas in
accordance with the Commission's rules. On May 28, 1997, the Common Carrier Bureau
("Bureau") issued an Order modifying the Common Line average schedule formula (Common
Line formula) and approving the other average schedule formulas submitted by NECA. On
June 20, 1997, the National Exchange Carrier Association, Inc. ("NECA") filed a Petition for
Reconsideration ("Petition") regarding the Bureau's decision to modify the Common Line
formula. Six parties filed comments. On August 15, 1997, NECA filed reply comments. In
this Order, we affirm our decision directing NECA to modify the Common Line formula, as
specified in our May Order. Our decision helps to ensure that carriers serving the smallest
exchanges receive adequate settlements to cover the higher projected costs of serving those
exchanges, and that as exchanges become bigger, settlements per line are reduced to reflect scale
economies.
2. On November 3, 1997, NECA filed proposed further modifications to the average
schedule formulas. On November 5, 1997, the Division issued a public notice soliciting
comments on NECA's filing. Three parties filed comments. On December 10, 1997, NECA
filed sixteen pages of errata to their November 3 filing. For the reasons discussed below we
approve in part and deny in part NECA's proposed further modifications to the average schedule
formulas.
II. BACKGROUND
3. Incumbent local exchange carriers ("incumbent LECs") that participate in the tariff
and pooling arrangements administered by NECA recover costs from the pools calculated either
on the basis of cost studies reflecting the carrier's own booked costs of service or on the basis
of average schedule formulas that estimate the cost of service. Prior to the adoption of the
Commission's access charge rules in 1984, incumbent LEC compensation
arrangements were handled through private contractual agreements within the telephone industry.
The industry's settlement mechanism based the amount of incumbent LEC compensation either
on cost studies or average schedule formulas that were used to estimate an incumbent LEC's cost
of service. To facilitate implementation of its access charge rules, the Commission incorporated
a modified version of the industry's existing average schedule arrangement. Average schedule
companies are defined in the Commission's rules as telephone companies that participated in
average schedule settlements on December 1, 1982.
4. An incumbent LEC participating in the NECA Common Line pool charges
interstate access rates according to a NECA tariff. The settlement process determines whether
the company has collected from its customers an amount greater than or less than its interstate
costs. If a company's interstate access charge revenues exceed its costs, including a reasonable
rate of return on its interstate investment, that company pays the difference to the NECA pools
in the settlement process. Conversely, if the company receives less in interstate access revenues
than its actual or estimated costs of service, the NECA pool pays the difference to the company.
"Cost companies" settle with NECA on the basis of their actual interstate costs of service.
"Average schedule companies" use formulas to estimate the average costs of service and settle
with NECA on the basis of those estimated costs. The average schedule formulas are designed
to simulate the disbursements that would be received by cost companies that are representative
of average schedule companies.
5. In developing interstate settlement formulas, NECA uses data from a sample of
both average schedule companies and cost companies. From average schedule companies,
NECA collects data that represents total investments and expenses. From cost companies,
NECA collects data concerning investment and expenses that are allocated to the interstate
jurisdiction and to access charge categories pursuant to the Commission's jurisdictional
separations and access charge rules. Average schedule companies do not have this type of data
because they are presumed to lack the resources to justify a requirement that they perform
separations and access charge cost studies to determine their compensation from interstate
services. Because data filed by average schedule companies are insufficient to calculate their
interstate revenue requirements, NECA develops simulated revenue requirements for the
sample average schedule companies based on analysis of the jurisdictional separations and access
element factors of a sample of cost companies. NECA then uses the simulated revenue
requirements of the sample average schedule companies to develop formulas that determine
settlements for all average schedule companies.
6. Section 69.606(b) of the Commission's rules requires NECA either to file revised
average schedule formulas on or before December 31 of each annual period, or to certify that
no such revisions are necessary. If the proposed settlement formulas are approved or modified
by the Commission, they take effect on July 1. On December 31, 1996, NECA filed proposed
revisions to the following average schedule formulas: Traffic Sensitive Central Office; Line Haul
Non-Distance Sensitive; Line Haul Distance Sensitive; Intertoll Dial Switching; Signal System
7; Special Access; Equal Access; Database Query and Interim Translation; Universal Service;
and Common Line. In its May Order, the Bureau approved all of the proposed formula changes
with the exception of the Common Line formula, which it modified.
7. NECA's proposed Common Line formula calculated an average schedule
company's Common Line settlement on the basis of its average number of access lines per
exchange. NECA used a linear spline model to determine the settlement per access line. A
linear spline is graphically represented by two straight lines joined at an inflection point. A
comparison of the NECA linear spline model with the actual cost data NECA obtained from the
sample, raised several concerns. For example, we questioned the appropriateness of the location
of the inflection point, which was based on 1985 data. We were also concerned that, using
the linear spline model, the settlement amounts for a large number of companies did not provide
an accurate simulation of the settlements that comparable cost companies would receive. We
determined that a reciprocal model would best fit the sample data, and, accordingly, we modified
NECA's proposed Common Line formula to incorporate a reciprocal model.
8. In deciding to modify the Common Line formula, we relied on two primary
methods of evaluation. First, we compared NECA's proposed formula with alternative models
using the R2 statistic as a measure of goodness of fit. Second, we compared the sum of
estimated settlements using NECA's model with the sum of the sample revenue requirements.
The modified formula was superior to NECA's formula under both criteria. Overall, the
Bureau's modified formula resulted in higher settlements to the smallest average schedule
companies and decreased total settlements by $2 million, compared with last year's settlements.
These results are discussed in greater detail below.
9. On June 16, 1997 NECA filed its annual tariffs and average schedule formulas
that reflected the Bureau's modification to the Common Line formula. On June 20, 1997 NECA
filed a petition seeking reconsideration of the Bureau's decision directing NECA to use the
Bureau's modified Common Line formula instead of NECA's proposed formula. On July 1,
1997 NECA's June 16 tariffs took effect.
III. POSITIONS OF PARTIES
A. NECA's Petition for Reconsideration
10. In its Petition for Reconsideration, NECA states that the Bureau's modification
to the Common Line formula contains significant statistical flaws. Specifically, NECA
maintains that the Bureau's modified formula produces substantial, unpredictable settlement
changes for average schedule companies, is overly sensitive to data from companies with small
numbers of lines per exchange, and understates the aggregate Common Line revenue
requirement for average schedule companies. In addition, NECA argues that the Bureau used
two erroneous data points in its calculations that should have been excluded.
11. NECA maintains that these flaws cause the modified formula to produce
settlements which fail to simulate cost company disbursements, as required by Section 69.606(a)
of the Commission's rules. In addition, NECA states that the modified formula would cause
serious fluctuations in settlements for average schedule companies. NECA also claims that the
Bureau failed to give notice of its proposed modified formula in violation of section 205 of the
Communication's Act and section 553 of the Administrative Procedure Act.
12. NECA requests that the Commission permit NECA to use its proposed linear
spline model to calculate the Common Line formula for 1997. NECA maintains that its linear
spline model more accurately targets population revenue requirements and will produce less
volatile, more predictable changes in settlements.
B. Comments and Reply Comments
13. Alliance supports NECA's petition, observing that the reciprocal model does not
posit a plausible relationship between costs and model variables. For instance, Alliance states
that when calculating the reciprocal model at low access lines per exchange levels, small
decreases in average lines per exchange produce large increases in the estimates of the cost per
line. Park Region, supporting NECA's petition, states that the Bureau did not consider the
effects that changes in the Common Line formula can have on individual average schedule
companies and the average schedule process in general. NTCA also supports NECA's
petition, stating that the Bureau's modification in the Common Line formula would result in
unpredictable fluctuations in year-to-year settlement rates. USTA, OPASTCO, NTCA, Park
Region, and Alliance emphasize that the Bureau should have sought comment on its proposed
modification.
14. MCI opposes NECA's petition. MCI states that, under NECA's formula, only
companies with higher cost structures than those subject to cost-based regulation will migrate
to cost company status. In addition, MCI states that: 1) removal of the two data points, which
NECA claims to be erroneous, has only a marginal effect on the intercepts, slopes, and the
coefficients of determination, or R2 values, of the models; 2) use of the Bureau's modified
formula does not necessarily increase rate instability because future formulas should better
address the Bureau's concerns; 3) NECA fails to justify the use of an inflection point derived
from 1985 data; and 4) NECA fails to show that its model minimizes the difference between
simulated revenue requirements for sample firms and estimated settlements for these firms.
15. In its reply comments, NECA included reports of two consultants, each of whom
conducted an analysis comparing a number of attributes of NECA's proposed linear spline model
with the reciprocal model. NECA's consultants claim that the reciprocal model is excessively
sensitive to companies with low numbers of lines per exchange. This sensitivity, they argue,
could lead to large year-to-year fluctuations in settlements.
C. NECA's Petition for Further Modifications
16. On November 3, 1997, NECA proposed interim modifications to several of the
average schedule formulas, that are scheduled to take effect on January 1, 1998 and remain in
effect through June 30, 1998. NECA asserts that these modifications comply with the rule
changes implemented in the Commission's universal service and access reform proceedings.
In addition to changes necessary to implement recent Commission orders, NECA proposes
substantial changes to the Common Line and its USF Loop Cost formulas. NECA also included
revisions to its sample data in its filing.
17. OPASTCO, NTCA and the USTA filed comments. These associations generally
support NECA's proposed revisions to the average schedule formulas.
IV. DISCUSSION
A. Petition for Reconsideration
18. The average schedule process is a method of modeling or projecting the costs of
providing interstate service for carriers that are so small that they are presumed to lack sufficient
financial resources or expertise to justify a requirement that they perform jurisdictionally
separated cost studies for determining their compensation in originating and terminating interstate
telecommunications services. For these companies, the costs of performing a cost study would
likely outweigh the potential benefit of a more precise cost estimate. To estimate the costs of
service, the average schedule companies use formulas that are designed to produce disbursements
that would be received by companies whose cost characteristics are representative of average
schedule companies. Both NECA and the Bureau have derived average schedule formulas by
applying regression analysis to simulated cost and other operational data for a sample of average
schedule companies. Our prior analysis of NECA's linear spline model for the Common Line
formula showed that it consistently and significantly overstated the average schedule revenue
requirement. We also found that several statistical models, including the Bureau's model, not
only reduce that tendency, but also more closely fit the data used to develop the formula. In
comparing various models, econometric regression is used to assess the relationship between cost
and other variables such as the number of access lines and the number of exchanges. We refer
to this measure of accuracy as "goodness of fit." In general, a model is considered more
accurate if it more closely tracks or follows these relationships. After careful evaluation of
alternative formulas, we required NECA to modify the model to match more closely the average
schedule revenue requirement based on the data in NECA's sample.
19. NECA's Model. We affirm our decision to reject NECA's model for the 1997
Common Line formula. NECA's regression analyses result in biased formulas that yield higher
settlements than are justified by the data. Table 1 lists in one column the simulated aggregate
Common Line revenue requirements that NECA computed for the average schedule companies
included in its sample for the past five years. The adjacent column in Table 1 then lists the
Common Line revenue requirements for those companies for each of those years that the use of
NECA's 1997 Common Line formula produces. As Table 1 shows, use of the NECA formulas
consistently results in settlements to the sample average schedule companies that exceed the
simulated revenue requirements computed by NECA. The overstatements range from a low of
4.9 percent to a high of 8.7 percent.
Table 1
Comparison of NECA's Estimated Total Costs with Total Costs of Sample
Year
Number of
Average
Schedule
Companies in
the Sample
NECA's
Simulated
Revenue
Requirements
for Sample
Companies
(Millions)
Sample
Company
Settlements
using NECA's
Proposed
Formula
(Millions)
Overstatement of Costs
Amount
(Millions)
Percent
1992
165
$ 113.7
$ 120.9
$ 7.2
6.3
1993
167
$ 133.8
$ 140.4
$ 6.6
4.9
1994
190
$ 122.7
$ 129.1
$ 6.4
5.2
1995
222
$ 138.8
$ 150.9
$ 12.1
8.7
1996
180
$ 103.0
$ 108.6
$ 5.6
5.4
20. In addition to the bias inherent in the results of NECA's average schedule
formulas during the past five years, we have determined that NECA's linear spline model does
not accurately reflect the sample data. First, these data do not demonstrate a sharp shift in
slope that would support the selection of an abrupt inflection point characteristic of NECA's
linear spline model. Second, even if NECA's use of a linear spline model were reasonable,
NECA's choice of an inflection point is not adequately supported by current data. NECA's
failure to recalculate the inflection point based on more current data is likely to be a primary
reason why NECA's linear spline model fails to accurately represent 1996 data.
21. Impact on Small Companies. We also affirm our decision to use a reciprocal
model to compute the Common Line formula. Upon examination of the data, we found several
different models, including the reciprocal model, that fit the data more closely than NECA's
proposal. At the outset, we observe that any modification of the Common Line formula will
change the amount of revenue received by individual companies. Such modifications are
warranted, however, to ensure that the formulas used by NECA appropriately compensate
average schedule companies. Park Region argues that the Bureau did not adequately consider
the effect that the use of the reciprocal model would have on the settlements of individual
average schedule companies. The Bureau staff did, in fact, analyze the effect that the modified
formula would have on every average schedule company. We found that approximately half the
600 average schedule companies would have a higher settlement under the modification than they
would using NECA's proposal. In fact, 85 percent of the average schedule companies with
fewer than 1,000 lines per exchange (291 of 344 companies) would receive a higher settlement
using the modified formula than they would receive if NECA's proposed formula were used.
Of the companies receiving higher settlements using the modified formula, a high proportion
were the smaller average schedule companies. The modified formula, therefore, has a positive
impact on these smaller incumbent LECs.
22. Validity of the R2 statistic. Table 2 shows a comparison of the R2 statistics, which
measure goodness of fit, for various models. NECA's linear spline model had the lowest R2
value and therefore was the poorest fit among these alternative models. The reciprocal model,
which is the model selected by the Bureau, had the highest Rư value and therefore was the best
fit. NECA argues in its petition that the Rư statistic used by the Bureau as a measure of
goodness of fit is not valid when the residuals do not conform to a normal distribution. We
reject that argument. The passage from the Kmenta text cited by NECA to support its challenge
regarding R2 statistics does not discuss R2 analysis. In fact, the Kmenta text supports the
Bureau's use of linear regression analysis with least squares criterion, stating that such analyses
produce the most reliable unbiased estimators in the presence of non-normal residuals. NECA
also argues that, according to the Scheffe text, mathematically correct statistical inferences rely
on the condition that residuals are distributed normally around zero. We reject that argument
as irrelevant because no statistical inferences were involved in our choice of the Rư statistic as
a measure of goodness of fit.
Table 2
Comparison of Common Line Settlement Models
ModelsFor a complete description of these models see For a description of the models see National Exchange Carrier Association, Inc., Memorandum Opinion
and Order, AAD 97-2, DA 97-1114, released May 28, 1997, at Attachment 2.
R2
NECA's Linear Spline
0.16
Bureau's Estimated Spline
0.19
Reciprocal
0.20
Log Reciprocal
0.17
23. Variance in Settlements over Time. NECA argues that, over time, use of the
Bureau's modification will result in excessive variation or volatility in the interstate average
schedule settlements for small average schedule companies. The basis for this volatility,
according to NECA, is that the reciprocal model is more sensitive to the variation for small
companies than the linear spline model. We find no basis for this argument. After careful
analysis of the changes in estimated settlements over the last five years using the reciprocal
model, we found no excessive variation for small average schedule companies. Changes in
technology and cost structure for average schedule companies, however, may require that we
consider other models over time. Our task is to ensure that, each year, the model used best
simulates cost companies based upon analysis of the cost company data. We do not believe that
selecting the optimal model for a given year will inevitably lead to excessive variation or
volatility. For the current year, we found that a reciprocal model fits the sample data supplied
by NECA better than the linear spline model filed by NECA. We do not expect significant
volatility in the cost data that form the basis for the Common Line average schedule model,
because these data are largely a function of embedded loop costs that do not typically increase
or decrease significantly from year to year.
24. Inclusion of Erroneous Data Points. In addition to challenging which model
more closely represents the data, NECA also argues that, in determining that the average
schedule model should be modified, the Bureau used two data points that resulted from
significant reporting errors. NECA argues that for one of the reported outliers it was unable to
allocate costs correctly to Common Line because the company had incorrectly included costs
from other operations. This reporting error should result in artificially inflated costs, implying
that inclusion of this data point would result in an average schedule that is too high. Exclusion
of this data point, however, results in an increased average schedule. For the other data point,
NECA claims that this point is not usable because the company operates in two states, but
reported data for only one state. Elsewhere, however, NECA reveals that it has complete
company information for that carrier, thus establishing that it could have determined the costs
of the company's operations in the second state. Furthermore, we agree with MCI's conclusion
that the faulty data points had only a marginal effect on the intercept slope and R2 values of the
Commission reciprocal model. Therefore, we conclude that both of these data points should
be included in the sample set used to derive the formula.
25. Settlements and Cost Trends. NECA argues that its formula would produce a 2.4
percent increase in per-unit Common Line settlements, while the Bureau's modification would
cause per-unit Common Line settlements to decline by 1.0 percent. According to NECA, this
one percent decline is inconsistent with universal service fund data, which show that Common
Line unit costs have increased each year from 1993 to 1996.
26. NECA's argument is misleading because it implies that the per-unit common line
settlements received by average schedule companies have accurately tracked the estimated per-
unit common line unit costs computed using simulated cost data. That has not, however, been
the case. As Table 1 shows, over the past several years the average schedule formulas have
produced results much higher than those derived using simulated cost data. This overstatement
was particularly pronounced for last year's Common Line formula settlements, which exceeded
the results from the sample by nearly nine percent. NECA, in citing a one percent decline
using the Bureau's modification, compared last year's Common Line formula settlements and
the level of settlements using the Bureau's modification. Given the significant overstatement of
last year's Common Line formula settlements, the alleged "one percent decline" does not
demonstrate that the Bureau's modification is unwarranted.
27. Sensitivity of the Model. NECA argues that our statistical analysis is unduly
sensitive to a few data points of the smallest companies. We disagree. In fact, as shown by the
Rư statistic, our analysis accurately reflects the sample data provided to the Commission by
NECA. Nonetheless, we believe that NECA has identified a potential area of concern regarding
the validity of the sample data upon which NECA bases its average schedule models.
Specifically, the sample data may not contain information for a sufficient number of very small
companies, and therefore the sample may not be an accurate representation of the entire group
of such companies. Accordingly, we hereby direct NECA to work with the Bureau staff to
develop a means of assuring that the sample data produced for design of future average schedule
models will accurately reflect all sizes of average schedule companies, including those with very
few lines. Should we determine that the use of sample data continues to be problematic,
however, the Bureau will consider requiring the use of expense and investment data from all
average schedule companies.
28. Notice. NECA and many commenters argue that the Bureau violated section 205
of the Communication's Act and section 553 of the Administrative Procedures Act by failing to
provide notice of the required modifications to NECA's proposed Common Line formula.
These assertions are without merit.
29. Section 553 of the Administrative Procedures Act is inapplicable because this
proceeding to consider NECA's average schedule formulas is not a rulemaking. Consideration
of the average schedule formulas is, instead, an intermediate step in the development of NECA's
interstate access tariff process. Under Section 69.606(a) of the Commission's rules, the
computation of average schedule payments is made in accordance with formulas approved or
modified by the Commission. Once the average schedule formulas are approved or modified,
they serve as a basis for NECA's computation of interstate access rates and subsequent filing
of interstate tariffs with the Commission, in accordance with sections 201 through 203 of the
Communications Act.
30. The tariffs are filed with the Commission prior to the date upon which they
become effective and, as required by section 61.38 of the Commission's rules, they must be
accompanied by all required cost support information, including the average schedule formulas.
Section 61.58 of the Commission's rules prescribes a notice period before the tariffs become
effective. During this notice period, all interested parties have an opportunity to review the
tariffs and the supporting documentation, including the average schedule formulas. Such parties
have an opportunity to petition the Commission to reject or to suspend and investigate any tariff
provision they believe to be unjustly discriminatory or otherwise unlawful. Once the tariff
becomes effective, parties may challenge the lawfulness of any tariff provision pursuant to
Section 208 of the Act by filing a complaint with the Commission.
31. In requiring modification of NECA's Common Line formula, we acted
consistently with section 69.606 of the Commission's rules governing average schedule formulas.
Subsequently, on June 16, 1997, NECA filed the interstate tariffs and the requisite cost support
information, including the modified average schedule formulas. We find that the notice period
following NECA's tariff filing provided full notice of the modifications to the average schedule
Common Line formula. Furthermore, we find that the tariff review process and the protection
of Section 208 of the Act afforded interested parties full opportunity to challenge the
modifications to that formula. We therefore reject as unfounded the assertion that no notice of
such modification was given.
B. Petition for Further Modification
32. NECA proposes minor modifications to the Traffic Sensitive Central Office,
Special Access, Signalling System 7, Intertoll Dial Switching, Line Haul-Distance Sensitive, and
Line-Haul Non-Distance Sensitive formulas. The effect of these changes on the revenue
requirement for averages schedule companies is minimal. We have reviewed NECA's proposed
changes to these formulas and find them consistent with the sample data and the Commission's
recent universal service and access reform orders. We therefore approve them. For the
following reasons, we do not approve NECA's proposed modifications to the Common Line or
USF loop cost formulas. NECA's proposed modifications of these formulas would increase the
average schedule companies' 1998 settlements by approximately $21 million.
33. Common Line Formula. We deny NECA's request to implement its proposed
Common Line formula. In our May Order, we demonstrated that the Bureau's reciprocal model
presented a better fit for the data than the linear spline model NECA developed for its Common
Line formula. In its interim proposal, NECA again proposes a linear spline model. In this
Order, we affirm that decision. We have analyzed NECA's proposed formula based on the new
sample data submitted by NECA. Our analysis shows that the Rư, or goodness of fit index,
obtained using NECA's proposed Common Line formula is less than the Rư obtained using the
reciprocal model. Because this higher Rư indicates a better goodness of fit, the Bureau's
reciprocal model more accurately represents carriers' costs. We therefore direct NECA to use
the following reciprocal model:
Settlement = $7.152430 x (Common Line access lines) + $725.252636 x (No. of
exchanges)
34. High cost support. NECA seeks an increase of 173 percent in the USF support
for average schedule companies in 1998, as compared to 1997, from $9.369 million to $25.607
million. NECA calculates this increase in the USF expense adjustment by changing the manner
in which the sample average schedule company data is expanded to the total average schedule
company population. However, in reviewing NECA's calculations, we find that NECA's
actual costs per loop for its sample companies are generally less than the costs per loop derived
from its proposed formula. According to individual company estimated costs per loop calculated
by NECA, we find that 128 of the 195 sampled companies, or about two-thirds, have estimated
costs per loop that are less than those derived from NECA's proposed formula. We expect any
calculation of costs per loop from NECA's average schedule formula to reasonably approximate
the actual costs per loop for the sampled companies. Furthermore, we find no justification for
a substantial increase in high cost support payments to average schedule companies that NECA's
proposed USF formula would produce. The Commission's policy of controlling growth in high
cost support have been clearly established. The Joint Board recommended, and the
Commission adopted, interim rules that limit the rate of growth of the USF to the rate of growth
in the total number of working loops nationwide. Nationwide annual growth in loops over the
past several years has been approximately 4 percent. As noted above, NECA seeks an increase
in support of approximately 173 percent. NECA has offered no persuasive justification for
treating growth in the support for average schedule companies differently than the Commission's
rules that limit growth in the total high cost fund. We therefore find the $16 million dollar
increase in high cost support payments to average schedule companies excessive and
unsubstantiated. For these reasons, we direct NECA to continue using the USF average
schedule formula that became effective July 1, 1997.
V. CONCLUSION
35. We find, based on the record developed in response to NECA's Petition for
Reconsideration, that NECA's linear spline model does not provide a more reliable basis for
establishing the average schedule Common Line formula than the reciprocal model reflecting the
modifications specified in the May Order. Specifically, we found fundamental flaws in NECA's
method of evaluating sample cost data, which resulted in overstatement of NECA's simulated
revenue requirements. The Bureau's reciprocal model, on the other hand, more accurately
reflects the sample data provided to the Commission by NECA. As a result, carriers serving
the smallest exchanges will receive adequate settlements to cover the higher projected costs of
serving those exchanges. Finally, we find that NECA's assertions that the Bureau violated
section 205 of the Communication's Act and section 553 of the Administrative Procedures Act
by failing to provide notice of the required modifications to NECA's proposed Common Line
formula are without merit. We conclude that the greater accuracy of the Bureau's model
reinforces our decision directing NECA to modify its proposed Common Line formula
calculations. We therefore affirm our prior Order.
36. In addition, we approve all of the average schedule formula adjustments proposed
by NECA in its November 3 filing except those for Common Line formula and the USF Loop
Cost formula. The formulas we herein approve shall become effective on January 1, 1998. We
do not approve NECA's proposed formula changes for the Common Line and USF Loop Cost
formulas. The USF Loop Cost formula that became effective on July 1, 1997 shall remain in
effect. NECA shall use the Common Line formula specified in paragraph 34 above to become
effective on January 1, 1998.
VI. ORDERING CLAUSES
37. Accordingly, IT IS ORDERED, pursuant to Sections 0.91, 0.291 and 69.606 of
the Commission's rules, 47 C.F.R 0.91, 0.291 and 69.606, 47 U.S.C. 154(i) and (j), 201,
202, 203, 205, 218 and 403, that NECA's Petition for Reconsideration IS DENIED.
38. IT IS FURTHER ORDERED, pursuant to Sections 0.91 and 0.291 of the
Commission rules, 47 C.F.R. 0.91, 9.291, that the average schedule formulas proposed by
the National Exchange Carriers Association, Inc. on November 3, 1997, for: Traffic Sensitive -
Central Office; Special Access; Signalling System 7; Intertoll Dial Switching; Line Haul -
Distance Sensitive; and Line-Haul Non-Distance Sensitive, SHALL BECOME EFFECTIVE
January 1, 1998.
39. IT IS FURTHER ORDERED, pursuant to Sections 0.91 and 0.291 of the
Commission rules, 47 C.F.R. 0.91, 9.291, that the average schedule formulas proposed by
the National Exchange Carrier Association, Inc. for Common Line and, USF Loop Costs, ARE
NOT APPROVED.
40. IT IS FURTHER ORDERED, pursuant to Sections 0.91 and 0.291 of the
Commission rules, 47 C.F.R. 0.91, 9.291, that the average schedule formula that became
effective on July 1, 1997 for USF Loop Costs, SHALL REMAIN IN EFFECT and that the
Common Line formula specified in this Order SHALL BECOME EFFECTIVE on January 1,
1998.
41. IT IS FURTHER ORDERED, pursuant to Section 4(i) of the Communications Act
of 1934, as Amended, 47 U.S.C. 154(i), and Sections 0.91 and 0.291 of the Commission's
rules, 47 C.F.R. 0.91, 0.291, that THIS ORDER IS EFFECTIVE UPON ITS RELEASE.
42. IT IS FURTHER ORDERED, that the National Exchange Carrier Association,
Inc., SHALL FILE revised tariffed rates in accordance with this Order. For these purposes we
will waiver Sections 61.58 and 61.59 of our rules, 47 C.F.R. 61.58, 61.59, to allow NECA
to file its revised rates on 2-days notice, on a non-streamlined basis, on December 30, 1997, to
become effective January 1, 1998, if it chooses to do so. NECA shall cite the DA number of
this Order for its permission.
FEDERAL COMMUNICATIONS COMMISSION
A. Richard Metzger, Jr.
Chief, Common Carrier Bureau
APPENDIX