5. Revisions to the High Cost Threshold Level

As a matter of high priority for reform, we turn to the issue of the whether, or in what manner, to revise the threshold level for "high cost" lines eligible for B-Fund support. As prescribed under current rules, only those lines served by a COLR in areas with costs above a designated "high cost" threshold qualify for
B-Fund support. The threshold governs the number of lines eligible for subsidy support. Only those primary residential lines in service areas in which the adopted proxy costs exceed the threshold qualify for B-Fund subsidy. To the extent that we can revise the threshold to focus more effectively on applying subsidy funds only to those lines that are located in genuinely high cost areas, the required level of B-Fund subsidies can be targeted more efficiently.

In D.96-10-066, we adopted a benchmark threshold equal to the higher of either (a) the statewide average cost of basic service (set at $20.30 per line), or (b) the basic flat rate plus End-User Common Line (EUCL) charge.⁶³ Because AT&T California's rate, including the EUCL, is $15.07 per line, its high cost threshold is $20.30 per line. AT&T draws additional subsidy on qualifying high cost lines equal to the difference between its $15.07 rate and the $20.30 benchmark. Verizon California's basic rate, including the EUCL is $23.75, which constitutes its threshold. The corresponding threshold for SureWest is $25.40. For Frontier, the threshold is $24.35. In this interim decision, we revise the threshold for high cost support funding to limit the number of lines qualifying for a subsidy to a more reasonable level, as discussed below.

5.1 Parties' Positions

Parties generally support revising the high cost threshold, but disagree about the extent and methodology by which to do so.⁶⁴ AT&T argues that the existing threshold formula based on system-average costs should continue to apply, and that $20.30 system-average figure should merely be updated to reflect more recent data.⁶⁵

Various parties argue that the existing threshold overstates the level of subsidy required to support universal service by ignoring revenues from services which are bundled with the residential access line in high cost areas. The ILECs and many intermodal competitors (e.g., cable and wireless) offer service bundles and features that include not only basic exchange service, but also broadband, wireless, caller ID, voicemail, and video services, among others. In the URF proceeding, the ILECs claimed that two-thirds or more of their customers subscribe to service bundles (the proportion depends in part on the definition of bundles). AT&T estimated that, as of July 2006, only 10.8% of its billed residential revenues were for basic service only without additional bundled services from AT&T or its affiliate.⁶⁶ Thus, for example, the current CHCF-B arrangement may provide a $10 per month subsidy to support a line which, in addition to basic service, generates an additional $40 per month for the ILEC and its affiliates from voice mail, a feature package, and DSL - none of which are allocated any part of the "high cost" line.

Since the COLR is receiving compensation through bundled services being offered to residential customers, various parties argue that the threshold should be revised to recognize such revenues. DRA argues that the ILECs may actually require no subsidy to provide lines in a particular high cost area where the full range of revenues that already provide full compensation to the ILECs and their affiliates for such lines. Time Warner argues that a conservative estimate for 2006 of the average revenue per line for bundled services for Verizon and AT&T California operations is $33.35.⁶⁷

Verizon advocates setting the B-Fund benchmark equal to the benchmark used by the FCC for evaluating funding requirements under the federal universal service program. The FCC applies this benchmark as the basis for each state to certify whether rural rates charged within the state are "reasonably comparable" to urban rates in order to qualify for federal universal service funds. For purposes of determining if such rates are "reasonably comparable," the FCC adopted a benchmark based on national urban residential rate data, set at two standard deviations above the average urban rate.

The FCC characterized this benchmark as establishing a "safe harbor," or presumption, that rates in high cost areas that do not exceed this benchmark are "reasonably comparable" to national urban rates. Verizon identified the safe harbor rate as $34.21 per line.⁶⁸ The FCC permitted states with rates below the benchmark to certify that their rates are "reasonably comparable" without requiring additional information, or to rebut the presumption by demonstrating that other factors beside basic service rates affect comparability.⁶⁹

Verizon argues that a necessary implication of resetting the B-Fund benchmark to the "safe harbor" level is that basic rate freeze would be lifted and subject to upward adjustment. Based on Verizon's logic, rate increases up to the level of the "safe harbor" rate would be considered "reasonably comparable" consistent with universal service goals.

Cox suggests that the Commission could consider use of the FCC's safe harbor rate as a starting point, and then evaluate the need for modifications to make the threshold more applicable to California circumstances. For example, the FCC rate could be adapted to reflect average urban rates for California, with the threshold calculated at two standard deviations above that average. SureWest recalculated the FCC "safe harbor" rate to be $26.43 per line, adjusted for California-specific rates for AT&T and Verizon. The B-Fund benchmark is based on the costs to provide basic service consistent with the Commission's Universal Service goals, whereas the FCC safe harbor rate includes other taxes and fees that do not relate to the provision of universal service. SureWest thus also recalculated the "safe harbor" rate to remove costs to customers that are not revenue to the carrier, resulting in a rate of $17.98 (incorporating California-specific rates).

Other parties (Sprint and Time-Warner) argue that the threshold should be based upon the level of expenditures that a customer residing in a high cost area can reasonably afford, consistent with the overall goal of a 95% penetration rate for basic service. Parties sponsoring this approach offered data from the U.S. Census and the FCC regarding the level of expenditures that consumers devote to basic telecommunications services. Based on demographic data regarding average household expenditure levels for telecommunications services, Sprint proposes the benchmark be raised to between $36 and $41 per line.

5.2 Discussion

We conclude that the current threshold is overly inclusive and allows subsidy support in areas where it is not needed to meet universal service goals. Moreover, even if system average cost was still an appropriate basis for a high cost benchmark threshold, the currently adopted figure for system average cost of $20.30 is extremely stale, and an updated cost study would be necessary to derive a more current figure. Reliance on the system average cost as a high cost benchmark is no longer appropriate, however, given current market and regulatory conditions. The criteria underlying the benchmark needs to be revised to reflect more accurately the goal of limiting subsidies to what is required to ensure that basic service remains affordable in high cost areas.

We should have looked at the B-Fund every three years, and the current benchmark and threshold data should have been updated pursuant to an earlier review. As articulated in D.96-10-066,⁷⁰ such periodic review was intended to ensure that the overall size of the Fund stayed within reason, subject to adjustment as competition and technology evolved. By conducting such period reviews, we anticipated that the need for ongoing high cost support may be reduced over time. The three-year review interval was expected also to provide time to determine whether new entrants were willing to serve high cost areas with the subsidies provided. We also anticipated that an auction mechanism could be a possible vehicle for subsequent determination of subsidy amounts instead of conducting resource-intensive updates of cost proxy inputs.

Since the periodic three-year review process has not been performed as originally intended, the task before us now is to move forward expeditiously with long-overdue reforms. Our priority in this first phase of the proceeding is to revise the benchmark threshold, as discussed below. Next, we shall address subsequent reform measures for Phase II of the proceeding.

By resetting the threshold to a more reasonable level as an initial reform, we will limit the number of lines that qualify for support and reduce the size of the B-Fund. As a result, we expect the balance in the B-Fund to decline by approximately 74%, assuming no change in the high cost proxy per line. After we complete the update of the high cost proxy per line, we will implement further revisions in the level of the Fund. Likewise, the B-Fund surcharge will be reduced to reflect a lower level of support payments.

The currently adopted B-Fund threshold levels do not effectively serve the purpose of limiting subsidies to areas that are truly "high cost." The current threshold levels are based on the premise that any cost in excess of the statewide average (presumed to be $20.30 per line) constitutes "high cost." By providing subsidy support wherever costs exceed $20.30 or the flat rate plus EUCL (whichever is higher), the fund subsidizes prices in excess of what is required to meet universal service goals.

The benchmark should no longer be based upon system average costs, even assuming updated underlying data. A benchmark based upon a utility's system average costs is a poor surrogate to measure what amount a customer can reasonably afford to pay for basic service. The definition of "high cost" therefore should not be defined simply as anything above a system average figure. Instead, the benchmark should be revised based on a standard of affordability by customers rather than system average costs of the utility.

Merely updating the system average cost will not inform us concerning the level of cost support actually needed to keep basic rates affordable in high cost areas. The proper focus of a benchmark should be the affordability by the customer and reasonable comparability of rates between rural and urban areas. Universal service goals are attained when rates charged for basic service are affordable for up to at least 95% of customers within California.⁷¹ There is no justification to subsidize lines based on a designation as "high cost" where such costs are already within an affordable range.

The goal of universal service "affordability" does not necessarily require the exact same rates be charged to rural and urban customers, but is based on the principle of "reasonable comparability." As a standard of affordability in rural high cost areas, Section 254 of the Act requires access in rural and high cost areas based on a "reasonably comparable" standard, as follows:

Consumers in all regions of the Nation, including low-income consumers and those in rural, insular, and high cost areas, should have access to telecommunications and information services, including interexchange services and advanced telecommunications and information services, that are reasonably comparable to those services provided in urban areas and that are available at rates that are reasonably comparable to rates charged for similar services in urban areas.⁷²

The principle of "reasonable comparability" allows for more flexibility in deriving an appropriate high cost benchmark than is reflected in the current B-Fund methodology based on system average costs. Setting the B-Fund threshold level based on "reasonable comparability" will more effectively delineate truly high cost areas, and limit subsidy support only to areas where costs exceed a reasonably affordable rate level. We find such an approach to be more in step with the statute's "affordability" standard.

We decline to adopt the use of the FCC "safe harbor" rate as a basis for revising or indexing the B-Fund benchmark. While we find merit in such an approach, the B-Fund benchmark is intended to provide a demarcation of "high cost" regions eligible for B-Fund subsidies based upon affordability of California-specific costs. By contrast, the FCC "safe harbor" rate is based upon nationwide revenues and other charges (such as 911 fees and sales taxes).

Even though SureWest offered certain refinements to the FCC "safe harbor" rate to reflect more California-specific data, we still find the resulting figures are inappropriate as a basis to set a high cost threshold for B-Fund purposes. Even with these refinements, the resulting rate still represents a measure that is not directly relevant to the issue at hand, namely, affordability of basic service costs by customers.⁷³ Moreover, the figure as recalculated by SureWest is distorted by the use of regulated and capped rates, and is therefore not the best surrogate of actual average cost of service.

We likewise reject parties' proposals to set the revised high cost threshold based upon tracking of the average revenues that the ILEC recovers for packages that it markets on a bundled service basis. It would be improper to index the threshold based on a specific carrier's prices for other services marketed on a bundled basis in addition to basic service. We no longer regulate carriers' prices for such nonbasic services, and attempting to index the B-Fund threshold based on such prices would be an improper reversal of procompetitive policies adopted in URF in D.06-08-030. Moreover, proponents of this approach focus only on the revenue from other services but ignore offsetting costs of such services. Since we no longer apply cost-based regulation to such services, however, there would be no basis to quantify the costs for such services. Any attempt to quantify such costs would conflict with URF as adopted in D.06-08-030 which favors the competitive discipline of the marketplace rather than cost-of-service regulation. Therefore, there is no basis to set a threshold level indexed to the total specific revenues associated with a COLR's service bundle.

As a basis for revising the benchmark to reflect a more relevant measure of affordability and reasonable comparability with rates in urban areas, we conclude that the most appropriate criteria relates to customers' ability to afford basic service. For this purpose, we shall consider relevant demographic data regarding consumer expenditures on telecommunications services. In this regard, Time Warner/CCTA provided the results of the FCC's annual survey of residential monthly phone rates for flat-rate residential service (2006) which show a range from $16.01 per line in Anaheim to $25.38 per line in Long Beach.⁷⁴ The same survey reports substantially higher rates in other cities (including cities served by AT&T and Verizon) without any reported universal service concerns. For example, residential flat rates of $34.33 were reported in West Memphis, Arkansas and 33.82 in Racine, Wisconsin. Of the 95 cities surveyed by the FCC, AT&T's California rates were the lowest in the nation.⁷⁵

Sprint provided similar data from the FCC and the U.S. Census Bureau. Based on this data, Sprint argues that an average household could realistically spend well over $30 per month on telecommunications services. The FCC reports that: "About 2% of all consumers expenditures are devoted to telephone service. This percentage has remained virtually unchanged over the past 20 years, despite major changes in the telephone industry and telephone usage." The stability in the percentage of household expenditures devoted to telephone usage over the past 20 years provides a solid basis upon which to establish an affordability benchmark. These FCC and Census Bureau data sources reveal that the national average household expense for wireline local exchange service remained at about $36 per month between 2000 and 2005.⁷⁶

As an alternative measure of affordability, Sprint suggests that the high cost benchmark could be set equal to 50% of consumers' average expenditures for all telecommunications services (currently equal to approximately $82 per month).⁷⁷ Such an approach would produce a benchmark of $41 (=$82 * 50%).

We conclude that the bounds of what constitutes affordable basic service may cover a range of demographic data. For purposes of a high cost benchmark, however, we require a specific figure. We conclude that the figure of $36 per month, representing average household expenditures on basic service, provides a reasonable threshold benchmark and is within the range of affordability for basic service. Given the range of report data regarding household expenditures for telecommunications services as presented by Sprint, we consider the $36 figure to provide a conservative proxy of basic service costs that a consumer may reasonably afford. This figure is at the lower end of the range suggested by Sprint. We shall thus authorize the revised B-Fund "high cost" benchmark to be increased to $36 per line by July 1, 2009.

In comments on the Proposed Decision, certain parties argue that the $36 per line figure is not a valid basis for a benchmark because it does not reflect California-specific data and does not distinguish bundled prices from stand-alone basic service prices. They argue that by including services beyond just basic service, the $36 figure produces an apples-and-oranges comparison that is not suitable as a high-cost benchmark proxy for setting support levels for basic service.

We recognize that the $36 benchmark figure incorporates a broader range of local exchange and toll services and is not limited only to basic service. For the limited purpose of setting a high-cost benchmark, however, we conclude that the $36 figure is reasonable, and that data underlying the benchmark need not track exactly with basic service elements subject to B-Fund support.

Our goal in setting the benchmark at $36 is to delineate "high-cost" lines that are eligible for B-Fund support to a COLR within a reasonable range. For lines with a stated cost below the benchmark, the COLR will not receive B-Fund support. For this purpose, the $36 per-line figure serves as a reasonable proxy for delineating basic access lines for which high-cost B-Fund support will be provided. The $36 benchmark, however, is in no way intended to serve as a cap on basic rate levels, or as a determination that retail rates for basic service alone as high as $36 would be affordable. Likewise, this benchmark level does not indicate that we believe it is appropriate for basic service to rise to a level of $36 per line.

The $36 benchmark is within the range of reasonableness based upon relevant criteria of affordability and comparability with urban area rates. The $36 benchmark reflects broad and stable trends in consumer expenditures for telecommunications services which may but not always involve the bundling of multiple services along with the basic access line.⁷⁸ As such, we find that the $36 level is suitable as a high cost proxy for our limited purpose today.

In its comments on the Proposed Decision, AT&T, in particular, objects to the adoption of the $36 benchmark, claiming that it is not based on the "basic service" that the B-Fund supports, but incorporates spending on all local services as well as some long-distance toll services. AT&T claims that the adoption of a $36 benchmark will result in an "unfunded gap" in cost recovery equal to the difference between the current charge for basic service plus the EUCL (set at $15.07) and the benchmark of $36. (AT&T Comments at 18.)

AT&T's objections to the use of $36 benchmark are rooted in an outdated regulatory paradigm that ignore marketplace realities concerning how carriers package local exchange services and the pricing dynamics through which costs are recovered. By framing its arguments narrowly in terms of a myopic comparison of the $36 benchmark with a $15.07 basic rate (including EUCL) AT&T ignores the broader context in which basic service lines are marketed. As discussed in detail below in rejecting AT&T's arguments regarding its desire for revenue neutrality, AT&T has considerable flexibility under our URF regulatory regime to bundle a variety of features (e.g., voicemail, call forwarding, Caller ID, etc.) together with the primary basic residential line offered to its retail customers. Even though the primary line remains subject to regulatory price controls, AT&T has flexibility under URF to adjust its prices for additional services bundled with the basic line, constrained only by competitive forces. In applying the $36 benchmark, we appropriately take into account this broader context in which residential lines are marketed with the flexibility to bundle the basic line with additional features and to flexibility price those additional features.

By referencing the range of local exchange and toll services that residential customers typically purchase, the $36 benchmark properly limits the level of B-Fund subsidy. Even though the $36 figure represents broad national averages, there is no reason to conclude that California customers' expenditure patterns differ significantly from this average figure. Thus, the $36 figure represents a reasonable approximation of a residential customer's average expenditure level for local exchange telephone service. It is thus reasonable to rely on the $36 figure as a basis for benchmarking the level of expenditures that can be considered affordable, consistent with our universal goal of a 95% penetration rate.

In view of the opportunities available to the ILEC to recover costs through the competitive marketing of an array of services to residential customers, we reject AT&T's narrowly-construed claim that any "unfunded gap" will result from the adoption of the $36 benchmark. We also reject AT&T's attempt to ignore the pricing flexibility granted under the provisions of URF as a basis for assessing the reasonableness of the $36 benchmark. AT&T claims that by recognizing the pricing flexibility that it now enjoys pursuant to URF as a consideration in setting the benchmark at the $36 level, the Commission is returning to "the monopoly regime of `implicit subsidies of averaged rates, and services priced above cost to support services priced below cost' that the Commission repudiated long ago." (AT&T Comments at 19, citing D.96-10-066, mimeo. at 17.)

Contrary to AT&T's claim, the broad-based recognition of the revenue-generating opportunities available to the ILEC as authorized under URF provides an entirely proper context for establishing a benchmark limiting the level of B-Fund subsidies required to support the provision of basic service in high cost regions. By taking into account the fact that the ILEC has a wide range of revenue-generating opportunities under URF, we are in no way returning to a regime of cross-subsidies. AT&T's claims in this regard are rooted in an outdated cost of service or NRF paradigm that has been superseded with the adoption of URF, a regulatory regime that AT&T itself strongly supported.

As explained in further detail below in reference to AT&T's revenue neutrality arguments, we no longer regulate the retail prices that the ILEC charges for most services besides basic service. Moreover, even though we will retain certain pricing restrictions on the basic service element during a transition period, we do not prohibit the marketing of service bundles at flexible prices that include a primary residential line. The pricing of such service bundles provides the ILEC with broad-based opportunities to recover costs, including primary lines costs, on a bundled basis, without Commission tracking or allocation of service-specific revenues and costs. With the advent of flexible pricing and service bundling available under URF, the notions of cross-subsidization lose their earlier meaning as related to the more restrictive NRF-era framework that no longer exists. The adoption of a $36 benchmark in no way constitutes a return to cross-subsidization of prices.

As discussed infra, we do intend to continue to apply rate caps on ILEC basic service as an interim transition measure, subject to an orderly transition process for any desired increases before full rate flexibility takes effect. We shall solicit comments in the next phase of this proceeding concerning the precise magnitude of such rate cap increases and the duration of the phase-in period for implementing such increases that should apply for each COLR as a prelude to granting full pricing flexibility for basic service. We do not intend for this transitional rate caps to continue indefinitely. Our goal is to set a discrete time period and maximum rate cap during which any increases in basic rate caps can be implemented on a gradual basis. In this manner, the transition to full rate flexibility can be implemented in a manner that avoids the risk of sudden large rate increases.

Because each COLR's basic service rates are currently set at different levels, adjustments in the rate cap may need to be phased in differently for each COLR. For example, the basic rate plus EUCL currently in effect for AT&T is below the system average of $20.30 per line. For the other COLRs, the basic rate plus EUCL exceeds the $20.30 per line system average. The precise phase-in time period and adjustments to the rate cap applicable to each COLR shall be determined in the next phase of the proceeding, and may depend on the current level of a particular COLR's basic rate.

During the period after the rate freeze is lifted, but while rate caps remain in effect, the COLR will have the flexibility to continue to charge basic rates that are below the cap, but will be precluded from increasing basic rates above the cap. Once the rate cap has increased to an appropriate level (to be determined in the next phase of the proceeding), we shall discontinue basic rate restrictions and authorize full pricing flexibility for basic rates. Although the COLR will have full pricing flexibility, the rates that are charged for basic service will not be used to determine the applicable level of B-Fund support. Instead, support levels will be calculated based upon the $36 per line benchmark (or any subsequent revision in the benchmark). The amount of B-Fund support will be limited to the difference between the $36 benchmark and the applicable per-line cost above that benchmark. The COLR will not be entitled to recover additional support for shortfalls between recovery through the basic rate plus EUCL and the $36 benchmark.

During the transition period as the existing benchmark is gradually increased from $20.30 up to $36, we shall continue to apply the existing formula for B-Fund support levels. Under this formula, for lines with costs above the high-cost benchmark, the support level includes the difference between the basic rate plus EUCL and the benchmark. Under this formula, for example, AT&T is able to recover additional subsidy to compensate for a basic rate plus EUCL which is below the $20.30 threshold. As a transitional measure, however, we shall discontinue this additional component of subsidy once the $36 benchmark is fully phased in. At that point, the per-line support shall be limited only to the applicable costs that exceed the $36 benchmark.

The use of the $36 revised benchmark will limit subsidy payments only to those lines in Census Block Groups (CBG) with a basic service cost proxy in excess of $36 per month. We shall order that subsidy support be phased out for those lines whose cost is equal to or less than the revised benchmark based on the timetable discussed below. As an interim measure, the revised benchmark levels shall be applied against existing High Cost CBG proxies. Once we complete the updating of CBG high cost proxies in the next phase of this proceeding, as discussed below, we shall further revise the applicable per-line subsidy disbursements accordingly.

As a result of the revisions in the benchmark authorized in this order, a significant number of CBGs that were previously considered "high cost" will now be excluded in computing B-Fund draws. To facilitate Commission staff review and monitoring of B-Fund subsidy draws submitted by COLRs for payment subsequent to this order, we direct that any new claims for B-Fund support clearly identify the specific CBGs, and associated proxy costs, that have been eliminated and that are no longer eligible for B-Fund support due to revisions in the threshold benchmark. COLRs shall provide this documentation separately for each successive change in the benchmark level, as set forth in the Appendix Table 1 schedule of this order.

We also shall authorize a B-Fund surcharge of 0.5%, which represents a reduction from the current 1.3% surcharge. We believe that this reduction is appropriate given the diminishing levels of claims we anticipate going forward as a result of the revised threshold we are adopting. We direct the ILECs to file Tier 1 advice letters to implement the revised 0.5% surcharge, to become effective by January 1, 2008.⁷⁹

⁶³ In D.96-10-066, we applied the EUCL charge as an offset to the fund because it covers a large share of the interstate portion of the LECs' nontraffic-sensitive embedded loop costs.

⁶⁴ See, e.g., Cox Opening Comments at 10; CCTA Opening Comments at 9; Sprint Opening Comments at 2, and; Verizon Opening Comments at 10-11.

⁶⁵ AT&T Comments of 4/27/07 at 3.

⁶⁶ AT&T Response to DRA Data Request 1-19, part f.

⁶⁷ Source Time Warner Comments dated 4/27/07, p. 4, citing 2006 ARMIS 43-03 (Revenues) and ARMIS 43-01 (Billable Access Lines). Revenue is the total of: Basic Area Revenue (Row 5001), Other Basic Area Revenue (Row 5060), End User Revenue (Row 5081), and Switched Access Revenue (Row 5100).

⁶⁸ This benchmark adjusts annually, and Verizon proposes that the B-Fund benchmark should likewise adjust annually to match updates in the FCC safe harbor benchmark. (See In the Matter of Federal-State Joint Board on Universal Service, CC Docket
No. 96-45 (rel. Oct. 27, 2003) ¶ 41.

⁶⁹ The FCC Order states that: "...standard deviation analysis of the relevant cost data supports the determination that the cost benchmark rejected by the court does in fact provide an appropriate level of non-rural High Cost support.[footnote omitted]. Standard deviation analysis is a commonly used statistical analysis that measures the dispersion of data points from the mean of those data points. [footnote omitted]. Both the [FCC] and state commissions have employed standard deviation analysis as a statistical standard for determining parity or comparability [footnote omitted]." See FCC Order on Universal Service (FCC 03-249) released October 27, 2003, page 38, ¶ 62.

⁷⁰ See D.96-10-066, 68 CPUC 2d, 524, 632-633.

⁷¹ See D.95-07-050, p. 548.

⁷² 47 U.S.C. § 254(b)(3).

⁷³ The Tenth Circuit approved the methodology chosen by the FCC, but remanded the matter for a better justification. See Qwest Communications International Inc. v. Federal Communications Commission (10^th Cir. 2005) 398 F.3d 1222, 1237 (Qwest II). Since the Tenth Circuit's decision, the FCC has not dealt with the remanded justification. It is not clear when or if, it will do so, and what changes, if any, may occur.

⁷⁴ The flat-rate for residential service includes the subscriber line charge, surcharges and taxes. (See Reference Book of Rates, Price Indices and Household Expenditures for Telephone Service, Industry Analysis and Technology Division, Wireline Competition Bureau, FCC, 2006, Tale 1.3, as cited in Time Warner/CCTA comments dated April 27, 2007, pp. 4-5.)

⁷⁵ After the rate freeze expires, AT&T will have flexibility to raise its basic residential rate to cover costs, subject to the phase-in process to be addressed in the next phase of this proceeding, and within the constraints of the competitive marketplace.

⁷⁶ See Sprint Comments of 4/27/07, footnote 25, citing Table 3.2, Average Monthly Household Telecommunications Expenditures by Type of Provider (Average are for only those households billed for service) at 3-4, in Trends in Telephone Service, February 2007 (Trends Report), published by the Industry Analysis and Technology Division, Wireline Competition Bureau, Federal Communications Division.

⁷⁷ Id. Sprint Comments, citing Trends Report, at 3-1, Table 3-1, Household Expenditures for Telephone Service at 3-3, showing that average annual expenditures on telephone service increased to $990 per household by 2004, equal to $82.50 per household per month.

⁷⁸ In this regard, we disagree with AT&T's claim that the use of the $36 benchmark would be improper for the same reasons that we reject parties' proposals to derive a benchmark based upon inclusion of revenues from COLR services other than basic service that are marketed to customers as part of a bundle. That approach would have required the Commission to review and track specific rate and cost levels for services that are no longer subject to price regulation. By contrast, the use of the $36 benchmark is based on a broad data set that does not require any review or monitoring a specific COLR's revenues or costs for services that are no longer subject to price regulation by the Commission.

⁷⁹ Since the resolution implementing the 0.5% surcharge will merely be a ministerial act, such advice letters shall be categorized as Tier I.