________________________________________ ) In the Matter of ) ) Federal-State Joint Board on ) CC Docket No. 96-45 Universal Service ) ) ________________________________________)
REPLY COMMENTS OF AD HOC
TELECOMMUNICATIONS USERS COMMITTEE
Economic Consultants: James S. Blaszak
Kevin S. DiLallo
Susan M. Gately Levine, Blaszak, Block & Boothby
Economics and Technology, Inc. 1300 Connecticut Avenue, N.W.
One Washington Mall Suite 500
Boston, MA 02108-2617 Washington, D.C. 20036-1703
617-227-0900 202-223-4980
May 7, 1996
200.03\usrepcov
SUMMARY
In calculating the total universal service support requirement, the first step is to determine what services should be supported. The record in this proceeding to date does not support expanding those services beyond the five core services identified by the Commission in the NPRM. Other services that have been proposed, such as interexchange service, do not satisfy the criteria established in Section 254 of the Communications Act.
An affordability benchmark should be established on a geographically disaggregated (i.e., not national) level. Merely calculating the nationwide average rate for local service would not produce an adequately focused result. Several other factors, including penetration levels, household income, and a defined basket of services within a defined geographic area, should be factored into the calculation of an affordable rate.
For any given geographic "cost zone," universal service support should fund only the difference between the calculated affordable rate for the universal service package of services and the total service long-run incremental cost ("TSLRIC") of providing those services (assuming that the TSLRIC exceeds the affordable rate).
Eligible telephone subscribers should be permitted to apply universal service support to the purchase of local service from any service provider, using any viable technology, and should not be restricted to the services of the incumbent LEC.
In calculating the funding requirement, the TSLRIC of providing universal service should be considered instead of incumbents' embedded costs. Reference to embedded costs encourages economic inefficiencies, thwarts the development of competition, and represents an obsolete regulatory compromise that is no longer appropriate in a competitive market.
A proxy cost model, such as the Benchmark Cost Model ("BCM") proposed by the Joint Sponsors in CC Dkt. No. 80-268, is a preferable methodology for calculating the universal service funding requirement. The BCM that has been proposed requires a handful of adjustments, and the Joint Sponsors apparently are making the needed modifications to the model. The Commission should reserve judgment on the model until the modifications have been completed.
The methodology prepared by National Economic Research Associates and proposed in the BellSouth Comments is flawed for several reasons, and should not be adopted.
Finally, widespread support has been demonstrated for eliminating the Carrier Common Line charge and increasing the Subscriber Line charge ("SLC"). The record indicates that the SLC could be increased moderately without reducing telephone subscribership.
TABLE OF CONTENTS
SUMMARY ii
INTRODUCTION 1
I. MCI's Overall Approach Is Sound 1
II The Benchmark Cost Model, With Adjustment, Could Be
Used As A Starting Point For Determining the Size of the
Universal Funding Requirement 6
III The Benchmark Cost Model Need Some Adjustments 9
IV The Gordon/Taylor Proposal to Set the Initial Support Based
Upon the Incumbent's Embedded Costs is Economically
Unsound and Will Perpetuate LEC Inefficiencies Into a Future
Competitive Market 11
V The Comments Demonstrate Wide Support For Eliminating
The Carrier Common Line Charge and Increasing The
Subscriber Line Charge 15
CONCLUSION 17
THE
FEDERAL COMMUNICATIONS COMMISSION
WASHINGTON, D.C. 20554
) In the Matter of ) ) Federal-State Joint Board on ) CC Docket No. 96-45 Universal Service ) )
REPLY COMMENTS
The Ad Hoc Telecommunications Users Committee (the "Ad Hoc Committee" or "Committee") hereby submits its Reply Comments in response to the Notice of Proposed Rulemaking and Order Establishing Joint Board, FCC 96-93, released March 8, 1996 ("NPRM").
INTRODUCTION
In an effort to distill the reams of comments filed by over 100 parties, the Committee has kept these Reply Comments brief. In the Committee's view, some of the most significant ideas (both positively and negatively) expressed in the initial comment round, which either merit reiteration or require refutation, include the following:
I. MCI's Overall Approach Is Sound.
In its initial Comments, MCI has proposed an approach for calculating the universal service funding requirement, allocating contribution responsibility, and distributing support where needed. Several aspects of MCI's proposal merit the Commission's careful consideration.
The goal of whatever procedure is adopted should be to replace all implicit and economically inefficient subsidy mechanisms with a simplified, explicit, and economically rational funding mechanism.[1] For example, the DEMS weighting subsidy, which has outlived its useful life, the Long-Term Support program, and the Carrier Common Line charge (which even LEC commenters such as BellSouth and Southwestern Bell Telecommunications admit is an implicit subsidy)[2] should all be eliminated.[3]
As a first step, MCI correctly proposes that the Commission must determine what services are to be supported by universal service subsidies. MCI Comments at 3. In this regard, certain commenters (e.g., Bell Atlantic) have proposed that interexchange service be included within the core services to be supported by universal service.[4] At least one other commenter (e.g., the United States Telephone Association ("USTA")) has proposed that voice-grade single business lines be included.[5]
These services should not be included among the services that are eligible for universal service support, because they do not satisfy the criterion of Section 254(c)(1)(A) of the Communications Act that supported services be "essential to education, public health, or public safety." It is difficult to conceive how interexchange services[6] and voice-grade single business lines satisfy this requirement.
Once the list of core services to be supported has been compiled, the MCI approach would require that a nationwide average monthly charge for "basic telephony" be established. MCI Comments at 4. While MCI's procedure is correct, its calculation of an "affordability" benchmark rate by reference to the current nationwide average local service rate may understate the affordability benchmark. For example, subscribership levels in areas with above-average local service rates are a useful and relevant indicator that affordability is not necessarily limited to the nationwide average rate. The record does not demonstrate that local subscribership levels are lower in areas where local service rates are above the national average. The Committee does not concur with MCI's statement that "the current nationwide average local service rate represents the most defensible definition of an "affordable" rate."[7]
In its initial Comments, the Committee articulated a number of factors that should be considered in determining what an "affordable" rate for universal service should be in lieu of simply calculating a nationwide average rate.[8] Such factors would include consideration of a defined basket of services within a defined geographic area, subscribership (or "penetration") rates, and household income. In short, a more focused view (i.e., on a more geographically disaggregated basis than nationally) of what constitutes an affordable rate is likely to yield a rate higher than the nationwide affordability benchmark rate proposed by MCI.
In any event, the Committee agrees with the procedural requirement that this rate be calculated as the second step toward calculating the universal service support payment obligation and receipt entitlement of eligible carriers.
The third step in the process, as MCI correctly points out, is to determine the total amount of the subsidy needed by using a cost proxy model such as the Benchmark Cost Model ("BCM") submitted by Joint Sponsors[9] on September 12, 1995 in CC Docket No. 80-268. The BCM is used to calculate the total service long-run incremental cost ("TSLRIC") of providing the core services comprising universal service for a limited number of different geographic "cost zones." The use of a proxy cost model instead of embedded LEC investment (referred to by the LECs as their "actual costs") has become one of the most vigorously contested issues in this proceeding.
The next step in determining individual providers' universal service funding needs would be, for each "cost zone," to multiply the number of lines within the cost zone (e.g., all lines within a wire center) by the calculated "affordable rate" for universal service core services.[10] In cost zones where the calculated affordable rate, multiplied by the number of residential subscriber lines, is less than the TSLRIC for providing core services to the same number of lines, universal service support would be available. Such support would not be available in cost zones where the TSLRIC, multiplied by the number of residential lines, is less than the calculated affordable rate times the number of residential lines within the cost zone.
Once the calculations for each cost center are completed and an aggregate national universal service funding requirement is established, a neutral third party would be responsible for allocating responsibility for contributing to the universal service fund, policing contributions, and distributing support payments to eligible providers of universal service within each cost center, according to the criteria discussed in Section IV, below.
As far as allocating the responsibility to contribute to universal service support among carriers, the Ad Hoc Committee continues to urge the Commission and the Joint Board to adopt a value-added approach, as detailed in the Committee's initial comments filed April 12. Such an approach would consider gross interstate revenues, less payments to other carriers, an approach similar to that advocated by MCI.[11] Because the value added by resellers to the provision of universal service is de minimis, resellers as a class of providers should be exempted from universal service support contributions, consistent with Section 254(d) of the Act.
Finally, the Committee agrees in principle with MCI's proposal that eligible residential subscribers should be permitted to apply credits toward the purchase of universal service core services from any provider, using any available and viable technology, and should not be required to purchase local service from the incumbent LEC or any other carrier or class of carrier, where multiple sources of local service exist.
II. The Benchmark Cost Model, With Adjustments, Could Be Used As a Starting Point for Determining the Size of the Universal Funding Requirement.
As the Committee stated in its initial Comments, the use of a proxy cost model to calculate the marginal, forward-looking costs of providing the package of core services supported by the universal service support mechanism (and therefore to determine the size of the universal service support requirement) is preferable to a methodology that uses the incumbent LECs' embedded costs. There are several reasons for this conclusion.
First, any system that provides for recovery of embedded costs is economically inefficient and, whatever its justifications in a monopolistic environment, is not justified in a competitive environment. As the Commission recently noted,[12]
[e]conomists generally agree that prices based on LRIC reflect the true economic cost of a service and give appropriate signals to producers and consumers and ensure efficient entry and utilization of the telecommunications infrastructure.
Second, competition would be harmed by perpetuation of a system in which incumbent LECs are indemnified for their past investment decisions, particularly where the parties contributing to the subsidies that provide the indemnification are potential competitors of the very LECs they are subsidizing.[13] Third, basing subsidy levels on LECs' embedded costs results in the least efficient LECs receiving the greatest support and therefore being handsomely rewarded for their inefficiency.
To the extent that incumbent LECs have argued that they are entitled because of their "regulatory social contract" to recover the costs of their investments in plant and equipment (see, e.g., Comments of Southwestern Bell Telephone Company at 23-24), such an argument should be flatly rejected. Whatever can be said for the former "regulatory social contract," the passage of the Telecommunications Act of 1996, at the fervent urging of the dominant LECs, has terminated it. Under the new regulatory contract, the LECs are permitted to enter markets previously closed to them but in return must open their former monopoly markets to competition. This includes eliminating barriers to entry including universal service support mechanisms that incorporate implicit subsidies for the incumbent LECs.
In this regard, the California Public Utilities Commission, in a 1995 universal service proceeding,[14] wrote:
The [New Regulatory Framework ("NRF")] decision in 1989 transformed the regulatory compact for Pacific and GTEC. The incentive-based regulatory framework was intended to expose shareholders to the risk associated with investments in order to provide an incentive to more efficient investment. . . .
NRF also signalled the gradual expansion of competition. . . . [I]nvestments made in anticipation of competition should not be regarded as stranded investment. . . .
Finally, there is no reason to believe that the incumbent LECs will have substantial stranded investments due to their past obligation to serve. According to their own representations, LECs' obligations are primarily due to serving high cost areas where competitors will be reluctant to enter. The assets associated with serving the high cost areas will not be stranded if the incumbent LEC continues to serve the high cost area, or it if resells its facilities to other providers.
In short, determination of the size of the universal service support requirement should not be based on a system that allows recovery of embedded costs or that rewards inefficient investment or operation by reference to LECs' embedded costs. The preferred alternative is a proxy cost model based on objective estimates of LECs' efficient incremental costs of providing local service. The Benchmark Cost Model ("BCM") proposed by the Joint Sponsors in CC Docket No. 80-268 would, with some adjustments, be an adequate device for calculating the size of the universal service support requirement.
III. The Benchmark Cost Model Needs Some Adjustments.
The Ad Hoc Committee generally supports the use of the BCM as a starting point for calculating the size of the universal service support requirement. The BCM is an engineering cost model that computes, by CBG, the cost of serving every area in the country except Alaska, for which necessary data is unavailable. As the National Cable Television Association ("NCTA") pointed out in its Comments,[15] however, the BCM should be adjusted somewhat before it is adopted.
For example, to capture the available economies of scale and scope, support should be computed on the basis of wire centers, rather than CBGs. Further adjustments should be made to reflect the fact that service subject to universal service support -- single-line service -- does not require the excess capacity that is embedded in the LECs' networks. The BCM also overstates costs, according to the BCM Study, because of its method for determining when to deploy fiber rather than copper; because the costs of digital subscriber loop do not reflect manufacturers' discounts; and because switch costs incorporated in the BCM do not reflect discounts routinely received by the LECs. In addition, the cost factor applied to the investment estimated by the BCM needs to reflect the forward-looking cost of providing residential local exchange service.
The Committee understands that the Joint Sponsors of the BCM are currently making adjustments to the BCM based on criticisms of the model, including many of those identified above. The Commission should reserve final judgment as to the usefulness of the BCM until such modifications are made and a revised BCM is proposed. Even in its present form, however, the BCM is superior to the methodology for calculating the universal service support requirement proposed by BellSouth.
IV. The Gordon/Taylor Proposal to Set the Initial Support Level Based Upon the Incumbent's Embedded Costs is Economically Unsound and Will Perpetuate LEC Inefficiencies Into a Future Competitive Market.
BellSouth Corporation and BellSouth Telecommunications (collectively, "BellSouth") submitted with their initial comments a paper prepared by Kenneth Gordon and William E. Taylor, of National Economic Research Associates ("NERA"),[16] which BellSouth proposes as a methodology for calculating the universal service support requirement. For the reasons set forth below, that model should be rejected in favor of a proxy cost model more closely resembling the BCM, as modified.
Gordon and Taylor (at 9-16) propose that the initial universal service support level be set at the difference between the incumbent's embedded cost and the rate level that is deemed "affordable." Under their scheme, the presence of at least one competitor whose LRIC is less than the incumbent's embedded cost for the basic service would be expected to bid down the market price below that being charged by the incumbent and, ultimately, induce the regulator to reduce the level of support so as to bring the net price back to the "affordable" level.
Gordon and Taylor have described a fanciful vision where regulators would periodically adjust and ultimately reduce the level of support in response to competitively-stimulated efficiency gains. They portray their initial embedded-cost-based support level as being only a starting point, calling for the invisible hand of the competitive marketplace to push regulators into making appropriate adjustments n the aggregate support requirement.
In proposing their scheme, Gordon and Taylor appear to assume that the support for below-cost pricing of basic service will somehow come from the ether rather than from the very same competitors who are supposed to help drive prices down. To Gordon and Taylor, universal service support would seem to be arriving exogenously on the scene like manna from heaven. In reality, whatever aggregate support is to be provided to basic universal service must necessarily be supplied by, among other things, the very competitors whom Gordon and Taylor count on to help bid down the overall support requirement. This key omission from their "numerical examples" conceals a fundamental flaw in the adjustment mechanism that they have described.
In their examples, Gordon and Taylor speak of the incumbent's embedded cost (at $25) and the competitor's LRIC at varying levels below that amount. They ignore the fact that the competitor will itself be required to make contribution to the universal service support fund in some rough relationship to its gross revenues, and that from the competitor's perspective, such payment obligations must be added to its LRIC in order to establish the floor price below which it cannot operate profitably. If the aggregate level of support per basic exchange access line is $10 based upon the incumbent's embedded cost, then new entrants will be saddled with a contribution requirement that will itself be driven by the incumbent's embedded cost.[17] Staying with the Gordon/Taylor hypothetical, suppose that the funding requirement is set at 20% of gross revenues.[18] In their Scenario 2, where the competitor's LRIC is $22, Gordon/Taylor suggest that the competitor will be able to undercut the incumbent's $15 price based upon receiving $10 if support. But suppose that the competitor sets its price at the $12 level posited by Gordon and Taylor. In that case, the competitor will be required to contribute another $3 [.20 x 12/(1-0.20)] to the universal service fund. Hence, from the competitor's perspective, the $22 LRIC translates into a $25 floor below which its price can not go. Thus, rather than adjusting out the excess of the incumbent's embedded cost over the most efficient producer's LRIC, the Gordon/Taylor approach would permanently lock in the incumbent's embedded cost through the funding system, and impose those costs upon all non-LEC providers.
That is not, as it turns out, the only flaw in these authors' scheme. They rely heavily upon the regulator to recognize the competitively-induced price decreases and to reduce the level of support so as to bring the price back to the affordability standard.[19] The adjustment mechanism they envision presupposes that the only form of competitive response will be a price change, rather than, for example, an expansion of the package of features that is included with the competitor's basic service offering. Perhaps this is because Gordon and Taylor would also restrict a competitor's ability to offer a basic service package that differs in an material respect from the product that is offered by the incumbent. For example, the competitor might offer its customers a different local calling area, or bundle in additional features like call waiting and call forwarding, rather than offer a monetary reduction in its price. How would the regulator, in such a case, recognize that a de facto price decrease had occurred? How would the regulator evaluate the extent of that decrease as a basis for adjusting the level of support? Gordon and Taylor do not even raise, let alone answer, these questions.
Setting the initial level of support on the basis of embedded cost serves only to protect incumbent carriers from competitive losses and to saddle their competitors with the burden of making the incumbent whole for past inefficiencies and poor business judgments. This approach serves only to reward the incumbents for past inefficiencies and imposes a permanent drag on the ability of competitors to enter and to effectively compete in the new market environment. There is no sound economic rationale for setting the support level on the basis of embedded costs, and the Commission should soundly reject this ill-conceived notion.
V. The Comments Demonstrate Wide Support for Eliminating the Carrier Common Line Charge and Increasing the Subscriber Line Charge.
A diverse range of parties, including some of the LEC commenters,[20] have observed that the Carrier Common Line charge ("CCLC") is an implicit subsidy that should be eliminated. SWBT has stated in this regard,[21]
Eliminating the interstate CCL and shifting recovery to end-users will lead to substantial economic gains for consumers as access price reductions generate toll reductions. Economists have measured efficiency losses attributable to the toll-to-local subsidy in the billions of dollars.
Moreover, a significant number of commenters have echoed the position the Ad Hoc Committee took in its Comments, that the Subscriber Line charge ("SLC") can be increased without adversely affecting subscribership levels.[22] Indeed, BellSouth has noted that telephone subscribership increased and the number of households without telephones decreased 19% between November, 1983 and November, 1989, despite the fact that SLCs were introduced in 1985 and slightly increased thereafter.[23]
One of the few dissenting voices on the issue of replacing the CCLC with an increased SLC was that of the National Association of Regulatory Utility Commissioners ("NARUC"). Although NARUC properly seems to recognize the economic inefficiency inherent in recovering non-traffic-sensitive costs through a per-minute, traffic sensitive charge,[24] it argued that increasing the SLC could reduce telephone subscribership levels. NARUC therefore argued that, "[i]f the Joint Board finds that it is not economically efficient to recover non-traffic sensitive NTS [sic] costs on a traffic sensitive basis via [the] CCL[C]," the per-minute CCLC should be replaced with a flat-rate charge paid by interexchange carriers ("IXCs") "because they use the LECs [sic] loop to provide their services."[25]
NARUC's proposal does not address two fundamental problems with even a flat-rate CCLC, namely, that it includes a subsidy flowing from users of interexchange services to users of local services and it requires potential competitors of the LECs to subsidize the incumbents. Unless these problems can be addressed, the CCLC should be eliminated and the SLC should be increased as necessary to enable the LECs to recover the proper interstate portion of subscriber loop costs.
CONCLUSION
For the foregoing reasons, the Ad Hoc Telecommunications Users Committee respectfully requests that the Commission take action in this docket consistent with the Committee's recommendations.
Respectfully submitted,
AD HOC TELECOMMUNICATIONS
USERS COMMITTEE
By: ____________________________
Economic consultants:
Susan M. Gately James S. Blaszak
Economics and Technology, Inc. Kevin S. DiLallo
One Washington Mall Levine, Blaszak, Block & Boothby
Boston, Massachusetts 02108-2617 1300 Connecticut Avenue, NW
(617) 227-0900 Suite 500
Washington, D.C. 20036
(202) 223-9019
May 7, 1996 Its Attorneys
I, Andrew Baer, hereby certify that true and correct copies of the foregoing Reply Comments of Ad Hoc Telecommunications Users Committee were served this 7th day of May, 1996 via U.S. first class mail, postage prepaid, upon the following parties: